8/22/24, 11:05 PM Project-ML-CODE

In [ ]: # Import essential libraries

import pandas as pd
import numpy as np

# Import TA-Lib for technical analysis

import talib as ta

# Import libraries for plotting and visualization

import matplotlib.pyplot as plt
import seaborn as sns

# Import libraries for machine learning preprocessing

from sklearn.pipeline import Pipeline
from sklearn.compose import ColumnTransformer
from sklearn.preprocessing import StandardScaler, MinMaxScaler

# Import machine learning models and utilities

from sklearn.ensemble import RandomForestClassifier
from sklearn.linear_model import LogisticRegression
from sklearn.svm import SVC
from sklearn.model_selection import GridSearchCV
from sklearn.model_selection import train_test_split
from sklearn.model_selection import cross_val_score
from sklearn.metrics import accuracy_score, f1_score, roc_auc_score
from sklearn.metrics import confusion_matrix, ConfusionMatrixDisplay
from sklearn.metrics import classification_report, precision_score, recall_score, r

# Import MiniSom for Self-Organizing Maps (SOM)

from minisom import MiniSom

# Import TensorFlow
import tensorflow as tf

# Latex and Markdown display

from IPython.display import display, Math, Markdown

WARNING:tensorflow:From c:\Users\bob19\anaconda3\Lib\site-packages\keras\src\losses.
py:2976: The name tf.losses.sparse_softmax_cross_entropy is deprecated. Please use t
f.compat.v1.losses.sparse_softmax_cross_entropy instead.

Step 1.1: Load the data from an Excel file

Load historical daily data of Guizhou Maotai from an Excel file.

In [ ]: # Load the historical daily data of Guizhou Maotai from an Excel file
file_path = 'data/600519.SH.xlsx'
data = pd.read_excel(file_path)

Step 1.2: Initial inspection of the data

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 1/132
8/22/24, 11:05 PM Project-ML-CODE

Inspect and preprocess the data, ensuring the date column is in the correct format and the
necessary columns are present.

In [ ]: # make sure the date column is in datetime format

data['date'] = pd.to_datetime(data['date'])

# Get the data from 2012 to 2019

start_date = '2012-01-01'
end_date = '2019-12-31'
data = data[(data['date'] >= start_date) & (data['date'] <= end_date)]

# sort the data by date in ascending order

data = data.sort_values('date')

# Ensure all column names are in lowercase

data.columns = data.columns.str.lower()

# Ensure all necessary columns are of correct data type

required_columns = ['open', 'high', 'low', 'close', 'volume']
for col in required_columns:
if col in data.columns:
data[col] = data[col].astype(float)

# Display the first few rows of the data to understand its structure
data.head()

Out[ ]: date open high close low volume

2508 2012-01-04 125.890804 126.725693 121.795244 121.617748 3387828.0

2509 2012-01-05 120.960354 121.749226 120.401570 119.882229 3012264.0

2510 2012-01-06 120.368700 123.622797 122.695872 119.415480 2519062.0

2511 2012-01-09 122.275141 123.649093 123.596501 119.494367 2616016.0

2512 2012-01-10 122.932534 127.968166 127.849835 122.347454 3923560.0

In [ ]: data.describe()

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 2/132

8/22/24, 11:05 PM Project-ML-CODE

Out[ ]: date open high close low volum

count 1945 1945.000000 1945.000000 1945.000000 1945.000000 1.945000e+0

2016-01-06
mean 343.831636 348.447328 344.355735 339.791978 3.950207e+0
06:56:04.935732736

2012-01-04
min 81.967920 83.057607 81.905849 81.388593 0.000000e+0
00:00:00

2014-01-09
25% 133.207590 135.213108 133.670677 131.527994 2.383883e+0
00:00:00

2016-01-06
50% 197.230775 202.281978 198.025872 193.899586 3.345317e+0
00:00:00

2018-01-02
75% 550.677366 561.690913 552.724050 546.088388 4.734770e+0
00:00:00

2019-12-31
max 1146.538136 1156.420158 1149.099451 1143.799856 2.891405e+0
00:00:00

std NaN 277.630109 280.924069 277.984950 274.538610 2.478505e+0

Step 1.3: Check for data types and missing

values
Ensure the data types are correct and check for any missing values in the dataset.

In [ ]: # Check the data types of each column

print(data.dtypes)

date datetime64[ns]
open float64
high float64
close float64
low float64
volume float64
dtype: object

In [ ]: # Check for missing values in the dataset

missing_values = data.isnull().sum()
print("Missing values in each column:")
print(missing_values)

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 3/132

8/22/24, 11:05 PM Project-ML-CODE

Missing values in each column:

date 0
open 0
high 0
close 0
low 0
volume 0
dtype: int64

Step 2.1: Feature engineering - Technical

indicators
Calculate various technical indicators such as Bollinger Bands, Moving Averages, and others
using TA-Lib.

In [ ]: # Define a list of time periods to be used for applicable indicators

time_periods_short = [5, 10, 14]
time_periods_medium = [20, 30, 50, 60]
time_periods_long = [100, 120, 200]

# Bollinger Bands
for period in time_periods_short + time_periods_medium:
data[f'bb_upper_{period}'], data[f'bb_middle_{period}'], data[f'bb_lower_{perio

# Double Exponential Moving Average (DEMA)

for period in time_periods_short + time_periods_medium:
data[f'dema_{period}'] = ta.DEMA(data['close'], timeperiod=period)

# Exponential Moving Average (EMA)

for period in time_periods_short + time_periods_medium + time_periods_long:
data[f'ema_{period}'] = ta.EMA(data['close'], timeperiod=period)

# Hilbert Transform - Instantaneous Trendline (HT_TRENDLINE)

data['ht_trendline'] = ta.HT_TRENDLINE(data['close'])

# Kaufman Adaptive Moving Average (KAMA)

for period in time_periods_short + time_periods_medium:
data[f'kama_{period}'] = ta.KAMA(data['close'], timeperiod=period)

# Moving Average (SMA)

for period in time_periods_short + time_periods_medium + time_periods_long:
data[f'sma_{period}'] = ta.SMA(data['close'], timeperiod=period)

# MESA Adaptive Moving Average (MAMA)

data['mama'], data['fama'] = ta.MAMA(data['close'])

# Moving Average with Variable Period (MAVP)

data['mavp'] = ta.MAVP(data['close'], periods=ta.SMA(data['close'], timeperiod=30))

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 4/132

8/22/24, 11:05 PM Project-ML-CODE

Step 2.2: Calculate the slope and slope

change for SMA
Compute the slope and the rate of change of the slope for selected Simple Moving Averages
(SMA).

In [ ]: # Example SMA columns (use actual SMA columns in your dataset)

sma_columns = ['sma_5', 'sma_10', 'sma_20', 'sma_60', 'sma_120', 'sma_200']

# Calculate the slope for each SMA over the past 5 days
for sma in sma_columns:
data[f'{sma}_slope'] = data[sma].diff(periods=5) / 5 # Change in SMA value ove

# Calculate the slope change (rate of change of slope) for each SMA
for sma in sma_columns:
data[f'{sma}_slope_change'] = data[f'{sma}_slope'].diff() # Day-over-day slope

Step 3.1: MidPoint and MidPrice indicators

Generate MidPoint and MidPrice indicators over specified periods.

In [ ]: # MidPoint over Period

for period in time_periods_short + time_periods_medium:
data[f'midpoint_{period}'] = ta.MIDPOINT(data['close'], timeperiod=period)

# MidPrice over Period

for period in time_periods_short + time_periods_medium:
data[f'midprice_{period}'] = ta.MIDPRICE(data['high'], data['low'], timeperiod=

Step 3.2: ADX, ADXR, APO, Aroon, and

Aroon Oscillator
Calculate various momentum and directional movement indicators.

In [ ]: # ADX and ADXR (Average Directional Movement Index and Rating)

for period in time_periods_short + time_periods_medium:
data[f'adx_{period}'] = ta.ADX(data['high'], data['low'], data['close'], timepe
data[f'adxr_{period}'] = ta.ADXR(data['high'], data['low'], data['close'], time

# APO (Absolute Price Oscillator)

for fastperiod in [5, 10, 12]:
for slowperiod in [26, 50]:
data[f'apo_{fastperiod}_{slowperiod}'] = ta.APO(data['close'], fastperiod=f

# Aroon and Aroon Oscillator

for period in time_periods_short + time_periods_medium:

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 5/132

8/22/24, 11:05 PM Project-ML-CODE

data[f'aroon_up_{period}'], data[f'aroon_down_{period}'] = ta.AROON(data['high'

data[f'aroonosc_{period}'] = ta.AROONOSC(data['high'], data['low'], timeperiod=

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 6/132

8/22/24, 11:05 PM Project-ML-CODE

C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\439499995.py:4: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'adxr_{period}'] = ta.ADXR(data['high'], data['low'], data['close'], timeper
iod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\439499995.py:3: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'adx_{period}'] = ta.ADX(data['high'], data['low'], data['close'], timeperio
d=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\439499995.py:4: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'adxr_{period}'] = ta.ADXR(data['high'], data['low'], data['close'], timeper
iod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\439499995.py:3: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'adx_{period}'] = ta.ADX(data['high'], data['low'], data['close'], timeperio
d=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\439499995.py:4: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'adxr_{period}'] = ta.ADXR(data['high'], data['low'], data['close'], timeper
iod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\439499995.py:9: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'apo_{fastperiod}_{slowperiod}'] = ta.APO(data['close'], fastperiod=fastperi
od, slowperiod=slowperiod)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\439499995.py:9: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'apo_{fastperiod}_{slowperiod}'] = ta.APO(data['close'], fastperiod=fastperi
od, slowperiod=slowperiod)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\439499995.py:9: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'apo_{fastperiod}_{slowperiod}'] = ta.APO(data['close'], fastperiod=fastperi
od, slowperiod=slowperiod)

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 7/132

8/22/24, 11:05 PM Project-ML-CODE

C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\439499995.py:9: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'apo_{fastperiod}_{slowperiod}'] = ta.APO(data['close'], fastperiod=fastperi
od, slowperiod=slowperiod)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\439499995.py:9: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'apo_{fastperiod}_{slowperiod}'] = ta.APO(data['close'], fastperiod=fastperi
od, slowperiod=slowperiod)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\439499995.py:9: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'apo_{fastperiod}_{slowperiod}'] = ta.APO(data['close'], fastperiod=fastperi
od, slowperiod=slowperiod)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\439499995.py:13: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'aroon_up_{period}'], data[f'aroon_down_{period}'] = ta.AROON(data['high'],
data['low'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\439499995.py:13: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'aroon_up_{period}'], data[f'aroon_down_{period}'] = ta.AROON(data['high'],
data['low'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\439499995.py:14: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'aroonosc_{period}'] = ta.AROONOSC(data['high'], data['low'], timeperiod=per
iod)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\439499995.py:13: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'aroon_up_{period}'], data[f'aroon_down_{period}'] = ta.AROON(data['high'],
data['low'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\439499995.py:13: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'aroon_up_{period}'], data[f'aroon_down_{period}'] = ta.AROON(data['high'],
data['low'], timeperiod=period)

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 8/132

8/22/24, 11:05 PM Project-ML-CODE

C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\439499995.py:14: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'aroonosc_{period}'] = ta.AROONOSC(data['high'], data['low'], timeperiod=per
iod)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\439499995.py:13: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'aroon_up_{period}'], data[f'aroon_down_{period}'] = ta.AROON(data['high'],
data['low'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\439499995.py:13: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'aroon_up_{period}'], data[f'aroon_down_{period}'] = ta.AROON(data['high'],
data['low'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\439499995.py:14: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'aroonosc_{period}'] = ta.AROONOSC(data['high'], data['low'], timeperiod=per
iod)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\439499995.py:13: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'aroon_up_{period}'], data[f'aroon_down_{period}'] = ta.AROON(data['high'],
data['low'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\439499995.py:13: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'aroon_up_{period}'], data[f'aroon_down_{period}'] = ta.AROON(data['high'],
data['low'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\439499995.py:14: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'aroonosc_{period}'] = ta.AROONOSC(data['high'], data['low'], timeperiod=per
iod)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\439499995.py:13: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'aroon_up_{period}'], data[f'aroon_down_{period}'] = ta.AROON(data['high'],
data['low'], timeperiod=period)

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 9/132

8/22/24, 11:05 PM Project-ML-CODE

C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\439499995.py:13: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'aroon_up_{period}'], data[f'aroon_down_{period}'] = ta.AROON(data['high'],
data['low'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\439499995.py:14: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'aroonosc_{period}'] = ta.AROONOSC(data['high'], data['low'], timeperiod=per
iod)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\439499995.py:13: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'aroon_up_{period}'], data[f'aroon_down_{period}'] = ta.AROON(data['high'],
data['low'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\439499995.py:13: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'aroon_up_{period}'], data[f'aroon_down_{period}'] = ta.AROON(data['high'],
data['low'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\439499995.py:14: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'aroonosc_{period}'] = ta.AROONOSC(data['high'], data['low'], timeperiod=per
iod)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\439499995.py:13: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'aroon_up_{period}'], data[f'aroon_down_{period}'] = ta.AROON(data['high'],
data['low'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\439499995.py:13: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'aroon_up_{period}'], data[f'aroon_down_{period}'] = ta.AROON(data['high'],
data['low'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\439499995.py:14: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'aroonosc_{period}'] = ta.AROONOSC(data['high'], data['low'], timeperiod=per
iod)

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 10/132

8/22/24, 11:05 PM Project-ML-CODE

Step 3.3: Other momentum indicators

Calculate additional momentum indicators including BOP, CCI, CMO, and others.

In [ ]: # BOP (Balance of Power)

data['bop'] = ta.BOP(data['open'], data['high'], data['low'], data['close'])

# CCI (Commodity Channel Index)

for period in time_periods_short + time_periods_medium:
data[f'cci_{period}'] = ta.CCI(data['high'], data['low'], data['close'], timepe

# CMO (Chande Momentum Oscillator)

for period in time_periods_short + time_periods_medium:
data[f'cmo_{period}'] = ta.CMO(data['close'], timeperiod=period)

# Momentum Indicators (MOM, ROC, ROCP, ROCR, ROCR100)

for period in time_periods_short + time_periods_medium:
data[f'mom_{period}'] = ta.MOM(data['close'], timeperiod=period)
data[f'roc_{period}'] = ta.ROC(data['close'], timeperiod=period)
data[f'rocp_{period}'] = ta.ROCP(data['close'], timeperiod=period)
data[f'rocr_{period}'] = ta.ROCR(data['close'], timeperiod=period)
data[f'rocr100_{period}'] = ta.ROCR100(data['close'], timeperiod=period)

# RSI (Relative Strength Index)

for period in time_periods_short + time_periods_medium:
data[f'rsi_{period}'] = ta.RSI(data['close'], timeperiod=period)

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 11/132

8/22/24, 11:05 PM Project-ML-CODE

C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:2: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data['bop'] = ta.BOP(data['open'], data['high'], data['low'], data['close'])
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:6: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'cci_{period}'] = ta.CCI(data['high'], data['low'], data['close'], timeperio
d=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:6: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'cci_{period}'] = ta.CCI(data['high'], data['low'], data['close'], timeperio
d=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:6: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'cci_{period}'] = ta.CCI(data['high'], data['low'], data['close'], timeperio
d=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:6: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'cci_{period}'] = ta.CCI(data['high'], data['low'], data['close'], timeperio
d=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:6: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'cci_{period}'] = ta.CCI(data['high'], data['low'], data['close'], timeperio
d=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:6: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'cci_{period}'] = ta.CCI(data['high'], data['low'], data['close'], timeperio
d=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:6: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'cci_{period}'] = ta.CCI(data['high'], data['low'], data['close'], timeperio
d=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:10: PerformanceWarni

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 12/132

8/22/24, 11:05 PM Project-ML-CODE

ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'cmo_{period}'] = ta.CMO(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:10: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'cmo_{period}'] = ta.CMO(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:10: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'cmo_{period}'] = ta.CMO(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:10: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'cmo_{period}'] = ta.CMO(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:10: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'cmo_{period}'] = ta.CMO(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:10: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'cmo_{period}'] = ta.CMO(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:10: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'cmo_{period}'] = ta.CMO(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:14: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'mom_{period}'] = ta.MOM(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:15: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'roc_{period}'] = ta.ROC(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:16: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 13/132

8/22/24, 11:05 PM Project-ML-CODE

using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram

e.copy()`
data[f'rocp_{period}'] = ta.ROCP(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:17: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'rocr_{period}'] = ta.ROCR(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:18: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'rocr100_{period}'] = ta.ROCR100(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:14: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'mom_{period}'] = ta.MOM(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:15: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'roc_{period}'] = ta.ROC(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:16: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'rocp_{period}'] = ta.ROCP(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:17: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'rocr_{period}'] = ta.ROCR(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:18: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'rocr100_{period}'] = ta.ROCR100(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:14: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'mom_{period}'] = ta.MOM(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:15: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 14/132

8/22/24, 11:05 PM Project-ML-CODE

data[f'roc_{period}'] = ta.ROC(data['close'], timeperiod=period)

C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:16: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'rocp_{period}'] = ta.ROCP(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:17: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'rocr_{period}'] = ta.ROCR(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:18: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'rocr100_{period}'] = ta.ROCR100(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:14: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'mom_{period}'] = ta.MOM(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:15: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'roc_{period}'] = ta.ROC(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:16: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'rocp_{period}'] = ta.ROCP(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:17: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'rocr_{period}'] = ta.ROCR(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:18: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'rocr100_{period}'] = ta.ROCR100(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:14: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'mom_{period}'] = ta.MOM(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:15: PerformanceWarni

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 15/132

8/22/24, 11:05 PM Project-ML-CODE

ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'roc_{period}'] = ta.ROC(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:16: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'rocp_{period}'] = ta.ROCP(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:17: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'rocr_{period}'] = ta.ROCR(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:18: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'rocr100_{period}'] = ta.ROCR100(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:14: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'mom_{period}'] = ta.MOM(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:15: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'roc_{period}'] = ta.ROC(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:16: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'rocp_{period}'] = ta.ROCP(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:17: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'rocr_{period}'] = ta.ROCR(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:18: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'rocr100_{period}'] = ta.ROCR100(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:14: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 16/132

8/22/24, 11:05 PM Project-ML-CODE

using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram

e.copy()`
data[f'mom_{period}'] = ta.MOM(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:15: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'roc_{period}'] = ta.ROC(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:16: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'rocp_{period}'] = ta.ROCP(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:17: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'rocr_{period}'] = ta.ROCR(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:18: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'rocr100_{period}'] = ta.ROCR100(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:22: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'rsi_{period}'] = ta.RSI(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:22: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'rsi_{period}'] = ta.RSI(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:22: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'rsi_{period}'] = ta.RSI(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:22: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'rsi_{period}'] = ta.RSI(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:22: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 17/132

8/22/24, 11:05 PM Project-ML-CODE

data[f'rsi_{period}'] = ta.RSI(data['close'], timeperiod=period)

C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:22: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'rsi_{period}'] = ta.RSI(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4266680648.py:22: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'rsi_{period}'] = ta.RSI(data['close'], timeperiod=period)

Step 3.4: Stochastic Oscillators and other

indicators
Generate stochastic oscillators, TRIX, ULTOSC, WILLR, and other indicators.

In [ ]: # Stochastic Oscillator (STOCH, STOCHF, STOCHRSI)

data['stoch_slowk'], data['stoch_slowd'] = ta.STOCH(data['high'], data['low'], data
data['stochf_fastk'], data['stochf_fastd'] = ta.STOCHF(data['high'], data['low'], d
for period in time_periods_short + time_periods_medium:
data[f'stochrsi_k_{period}'], data[f'stochrsi_d_{period}'] = ta.STOCHRSI(data['

# TRIX (1-day Rate-Of-Change of a Triple Smooth EMA)

for period in time_periods_short + time_periods_medium:
data[f'trix_{period}'] = ta.TRIX(data['close'], timeperiod=period)

# ULTOSC (Ultimate Oscillator)

data['ultosc'] = ta.ULTOSC(data['high'], data['low'], data['close'], timeperiod1=7,

# WILLR (Williams' %R)

for period in time_periods_short + time_periods_medium:
data[f'willr_{period}'] = ta.WILLR(data['high'], data['low'], data['close'], ti

# AD and ADOSC (Chaikin A/D Line and Oscillator)

data['ad'] = ta.AD(data['high'], data['low'], data['close'], data['volume'])
for fastperiod in [3, 6]:
for slowperiod in [10, 20]:
data[f'adosc_{fastperiod}_{slowperiod}'] = ta.ADOSC(data['high'], data['low

# OBV (On Balance Volume)

data['obv'] = ta.OBV(data['close'], data['volume'])

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 18/132

8/22/24, 11:05 PM Project-ML-CODE

C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3909819040.py:2: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data['stoch_slowk'], data['stoch_slowd'] = ta.STOCH(data['high'], data['low'], dat
a['close'])
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3909819040.py:2: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data['stoch_slowk'], data['stoch_slowd'] = ta.STOCH(data['high'], data['low'], dat
a['close'])
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3909819040.py:3: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data['stochf_fastk'], data['stochf_fastd'] = ta.STOCHF(data['high'], data['low'],
data['close'])
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3909819040.py:3: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data['stochf_fastk'], data['stochf_fastd'] = ta.STOCHF(data['high'], data['low'],
data['close'])
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3909819040.py:5: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'stochrsi_k_{period}'], data[f'stochrsi_d_{period}'] = ta.STOCHRSI(data['clo
se'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3909819040.py:5: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'stochrsi_k_{period}'], data[f'stochrsi_d_{period}'] = ta.STOCHRSI(data['clo
se'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3909819040.py:5: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'stochrsi_k_{period}'], data[f'stochrsi_d_{period}'] = ta.STOCHRSI(data['clo
se'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3909819040.py:5: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'stochrsi_k_{period}'], data[f'stochrsi_d_{period}'] = ta.STOCHRSI(data['clo
se'], timeperiod=period)

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 19/132

8/22/24, 11:05 PM Project-ML-CODE

C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3909819040.py:5: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'stochrsi_k_{period}'], data[f'stochrsi_d_{period}'] = ta.STOCHRSI(data['clo
se'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3909819040.py:5: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'stochrsi_k_{period}'], data[f'stochrsi_d_{period}'] = ta.STOCHRSI(data['clo
se'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3909819040.py:5: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'stochrsi_k_{period}'], data[f'stochrsi_d_{period}'] = ta.STOCHRSI(data['clo
se'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3909819040.py:5: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'stochrsi_k_{period}'], data[f'stochrsi_d_{period}'] = ta.STOCHRSI(data['clo
se'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3909819040.py:5: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'stochrsi_k_{period}'], data[f'stochrsi_d_{period}'] = ta.STOCHRSI(data['clo
se'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3909819040.py:5: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'stochrsi_k_{period}'], data[f'stochrsi_d_{period}'] = ta.STOCHRSI(data['clo
se'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3909819040.py:5: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'stochrsi_k_{period}'], data[f'stochrsi_d_{period}'] = ta.STOCHRSI(data['clo
se'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3909819040.py:5: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'stochrsi_k_{period}'], data[f'stochrsi_d_{period}'] = ta.STOCHRSI(data['clo
se'], timeperiod=period)

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 20/132

8/22/24, 11:05 PM Project-ML-CODE

C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3909819040.py:5: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'stochrsi_k_{period}'], data[f'stochrsi_d_{period}'] = ta.STOCHRSI(data['clo
se'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3909819040.py:5: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'stochrsi_k_{period}'], data[f'stochrsi_d_{period}'] = ta.STOCHRSI(data['clo
se'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3909819040.py:9: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'trix_{period}'] = ta.TRIX(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3909819040.py:9: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'trix_{period}'] = ta.TRIX(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3909819040.py:9: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'trix_{period}'] = ta.TRIX(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3909819040.py:9: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'trix_{period}'] = ta.TRIX(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3909819040.py:9: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'trix_{period}'] = ta.TRIX(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3909819040.py:9: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'trix_{period}'] = ta.TRIX(data['close'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3909819040.py:9: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'trix_{period}'] = ta.TRIX(data['close'], timeperiod=period)

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 21/132

8/22/24, 11:05 PM Project-ML-CODE

C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3909819040.py:12: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data['ultosc'] = ta.ULTOSC(data['high'], data['low'], data['close'], timeperiod1=
7, timeperiod2=14, timeperiod3=28)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3909819040.py:16: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'willr_{period}'] = ta.WILLR(data['high'], data['low'], data['close'], timep
eriod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3909819040.py:16: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'willr_{period}'] = ta.WILLR(data['high'], data['low'], data['close'], timep
eriod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3909819040.py:16: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'willr_{period}'] = ta.WILLR(data['high'], data['low'], data['close'], timep
eriod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3909819040.py:16: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'willr_{period}'] = ta.WILLR(data['high'], data['low'], data['close'], timep
eriod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3909819040.py:16: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'willr_{period}'] = ta.WILLR(data['high'], data['low'], data['close'], timep
eriod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3909819040.py:16: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'willr_{period}'] = ta.WILLR(data['high'], data['low'], data['close'], timep
eriod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3909819040.py:16: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'willr_{period}'] = ta.WILLR(data['high'], data['low'], data['close'], timep
eriod=period)

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 22/132

8/22/24, 11:05 PM Project-ML-CODE

C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3909819040.py:19: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data['ad'] = ta.AD(data['high'], data['low'], data['close'], data['volume'])
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3909819040.py:22: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'adosc_{fastperiod}_{slowperiod}'] = ta.ADOSC(data['high'], data['low'], dat
a['close'], data['volume'], fastperiod=fastperiod, slowperiod=slowperiod)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3909819040.py:22: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'adosc_{fastperiod}_{slowperiod}'] = ta.ADOSC(data['high'], data['low'], dat
a['close'], data['volume'], fastperiod=fastperiod, slowperiod=slowperiod)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3909819040.py:22: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'adosc_{fastperiod}_{slowperiod}'] = ta.ADOSC(data['high'], data['low'], dat
a['close'], data['volume'], fastperiod=fastperiod, slowperiod=slowperiod)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3909819040.py:22: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'adosc_{fastperiod}_{slowperiod}'] = ta.ADOSC(data['high'], data['low'], dat
a['close'], data['volume'], fastperiod=fastperiod, slowperiod=slowperiod)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3909819040.py:25: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data['obv'] = ta.OBV(data['close'], data['volume'])

In [ ]: # Hilbert Transform Indicators

data['ht_dcperiod'] = ta.HT_DCPERIOD(data['close'])
data['ht_dcphase'] = ta.HT_DCPHASE(data['close'])
data['ht_inphase'], data['ht_quadrature'] = ta.HT_PHASOR(data['close'])
data['ht_sine'], data['ht_leadsine'] = ta.HT_SINE(data['close'])
data['ht_trendmode'] = ta.HT_TRENDMODE(data['close'])

# MINUS_DI and PLUS_DI (Directional Indicators)

for period in time_periods_short + time_periods_medium:
data[f'minus_di_{period}'] = ta.MINUS_DI(data['high'], data['low'], data['close
data[f'plus_di_{period}'] = ta.PLUS_DI(data['high'], data['low'], data['close']

# MINUS_DM and PLUS_DM (Directional Movement)

for period in time_periods_short + time_periods_medium:

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 23/132

8/22/24, 11:05 PM Project-ML-CODE

data[f'minus_dm_{period}'] = ta.MINUS_DM(data['high'], data['low'], timeperiod=

data[f'plus_dm_{period}'] = ta.PLUS_DM(data['high'], data['low'], timeperiod=pe

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 24/132

8/22/24, 11:05 PM Project-ML-CODE

C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\2078805182.py:2: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data['ht_dcperiod'] = ta.HT_DCPERIOD(data['close'])
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\2078805182.py:3: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data['ht_dcphase'] = ta.HT_DCPHASE(data['close'])
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\2078805182.py:4: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data['ht_inphase'], data['ht_quadrature'] = ta.HT_PHASOR(data['close'])
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\2078805182.py:4: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data['ht_inphase'], data['ht_quadrature'] = ta.HT_PHASOR(data['close'])
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\2078805182.py:5: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data['ht_sine'], data['ht_leadsine'] = ta.HT_SINE(data['close'])
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\2078805182.py:5: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data['ht_sine'], data['ht_leadsine'] = ta.HT_SINE(data['close'])
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\2078805182.py:6: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data['ht_trendmode'] = ta.HT_TRENDMODE(data['close'])
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\2078805182.py:10: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'minus_di_{period}'] = ta.MINUS_DI(data['high'], data['low'], data['close'],
timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\2078805182.py:11: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'plus_di_{period}'] = ta.PLUS_DI(data['high'], data['low'], data['close'], t
imeperiod=period)

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 25/132

8/22/24, 11:05 PM Project-ML-CODE

C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\2078805182.py:10: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'minus_di_{period}'] = ta.MINUS_DI(data['high'], data['low'], data['close'],
timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\2078805182.py:11: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'plus_di_{period}'] = ta.PLUS_DI(data['high'], data['low'], data['close'], t
imeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\2078805182.py:10: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'minus_di_{period}'] = ta.MINUS_DI(data['high'], data['low'], data['close'],
timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\2078805182.py:11: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'plus_di_{period}'] = ta.PLUS_DI(data['high'], data['low'], data['close'], t
imeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\2078805182.py:10: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'minus_di_{period}'] = ta.MINUS_DI(data['high'], data['low'], data['close'],
timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\2078805182.py:11: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'plus_di_{period}'] = ta.PLUS_DI(data['high'], data['low'], data['close'], t
imeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\2078805182.py:10: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'minus_di_{period}'] = ta.MINUS_DI(data['high'], data['low'], data['close'],
timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\2078805182.py:11: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'plus_di_{period}'] = ta.PLUS_DI(data['high'], data['low'], data['close'], t
imeperiod=period)

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 26/132

8/22/24, 11:05 PM Project-ML-CODE

C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\2078805182.py:10: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'minus_di_{period}'] = ta.MINUS_DI(data['high'], data['low'], data['close'],
timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\2078805182.py:11: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'plus_di_{period}'] = ta.PLUS_DI(data['high'], data['low'], data['close'], t
imeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\2078805182.py:10: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'minus_di_{period}'] = ta.MINUS_DI(data['high'], data['low'], data['close'],
timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\2078805182.py:11: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'plus_di_{period}'] = ta.PLUS_DI(data['high'], data['low'], data['close'], t
imeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\2078805182.py:15: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'minus_dm_{period}'] = ta.MINUS_DM(data['high'], data['low'], timeperiod=per
iod)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\2078805182.py:16: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'plus_dm_{period}'] = ta.PLUS_DM(data['high'], data['low'], timeperiod=perio
d)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\2078805182.py:15: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'minus_dm_{period}'] = ta.MINUS_DM(data['high'], data['low'], timeperiod=per
iod)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\2078805182.py:16: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'plus_dm_{period}'] = ta.PLUS_DM(data['high'], data['low'], timeperiod=perio
d)

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 27/132

8/22/24, 11:05 PM Project-ML-CODE

C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\2078805182.py:15: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'minus_dm_{period}'] = ta.MINUS_DM(data['high'], data['low'], timeperiod=per
iod)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\2078805182.py:16: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'plus_dm_{period}'] = ta.PLUS_DM(data['high'], data['low'], timeperiod=perio
d)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\2078805182.py:15: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'minus_dm_{period}'] = ta.MINUS_DM(data['high'], data['low'], timeperiod=per
iod)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\2078805182.py:16: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'plus_dm_{period}'] = ta.PLUS_DM(data['high'], data['low'], timeperiod=perio
d)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\2078805182.py:15: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'minus_dm_{period}'] = ta.MINUS_DM(data['high'], data['low'], timeperiod=per
iod)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\2078805182.py:16: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'plus_dm_{period}'] = ta.PLUS_DM(data['high'], data['low'], timeperiod=perio
d)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\2078805182.py:15: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'minus_dm_{period}'] = ta.MINUS_DM(data['high'], data['low'], timeperiod=per
iod)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\2078805182.py:16: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'plus_dm_{period}'] = ta.PLUS_DM(data['high'], data['low'], timeperiod=perio
d)

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 28/132

8/22/24, 11:05 PM Project-ML-CODE

C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\2078805182.py:15: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'minus_dm_{period}'] = ta.MINUS_DM(data['high'], data['low'], timeperiod=per
iod)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\2078805182.py:16: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'plus_dm_{period}'] = ta.PLUS_DM(data['high'], data['low'], timeperiod=perio
d)

Step 4.1: Volatility indicators

Compute volatility indicators such as ATR, NATR, and TRANGE.

In [ ]: # Average True Range (ATR)

for period in time_periods_short + time_periods_medium:
data[f'atr_{period}'] = ta.ATR(data['high'], data['low'], data['close'], timepe

# Normalized Average True Range (NATR)

for period in time_periods_short + time_periods_medium:
data[f'natr_{period}'] = ta.NATR(data['high'], data['low'], data['close'], time

# True Range (TRANGE)

data['trange'] = ta.TRANGE(data['high'], data['low'], data['close'])

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 29/132

8/22/24, 11:05 PM Project-ML-CODE

C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3555839564.py:3: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'atr_{period}'] = ta.ATR(data['high'], data['low'], data['close'], timeperio
d=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3555839564.py:3: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'atr_{period}'] = ta.ATR(data['high'], data['low'], data['close'], timeperio
d=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3555839564.py:3: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'atr_{period}'] = ta.ATR(data['high'], data['low'], data['close'], timeperio
d=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3555839564.py:3: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'atr_{period}'] = ta.ATR(data['high'], data['low'], data['close'], timeperio
d=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3555839564.py:3: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'atr_{period}'] = ta.ATR(data['high'], data['low'], data['close'], timeperio
d=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3555839564.py:3: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'atr_{period}'] = ta.ATR(data['high'], data['low'], data['close'], timeperio
d=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3555839564.py:3: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'atr_{period}'] = ta.ATR(data['high'], data['low'], data['close'], timeperio
d=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3555839564.py:7: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'natr_{period}'] = ta.NATR(data['high'], data['low'], data['close'], timeper
iod=period)

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 30/132

8/22/24, 11:05 PM Project-ML-CODE

C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3555839564.py:7: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'natr_{period}'] = ta.NATR(data['high'], data['low'], data['close'], timeper
iod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3555839564.py:7: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'natr_{period}'] = ta.NATR(data['high'], data['low'], data['close'], timeper
iod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3555839564.py:7: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'natr_{period}'] = ta.NATR(data['high'], data['low'], data['close'], timeper
iod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3555839564.py:7: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'natr_{period}'] = ta.NATR(data['high'], data['low'], data['close'], timeper
iod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3555839564.py:7: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'natr_{period}'] = ta.NATR(data['high'], data['low'], data['close'], timeper
iod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3555839564.py:7: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'natr_{period}'] = ta.NATR(data['high'], data['low'], data['close'], timeper
iod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\3555839564.py:10: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data['trange'] = ta.TRANGE(data['high'], data['low'], data['close'])

Step 4.2: Ratio calculations

Calculate High-to-Low (H2L) and Open-to-Close (O2C) ratios, as well as daily returns.

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 31/132

8/22/24, 11:05 PM Project-ML-CODE

In [ ]: # Calculate High-to-Low (H2L) ratio

data['h2l'] = data['high'] / data['low']

# Calculate Open-to-Close (O2C) ratio

data['o2c'] = data['open'] / data['close']

# Calculate the daily return as the percentage change in the closing price
data['return'] = data['close'].pct_change()

C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4279441582.py:2: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data['h2l'] = data['high'] / data['low']
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4279441582.py:5: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data['o2c'] = data['open'] / data['close']
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4279441582.py:8: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data['return'] = data['close'].pct_change()

Step 4.3: Volume indicators

Calculate Money Flow Index (MFI) and Volume Rate of Change (VROC) indicators.

In [ ]: # Define a list of time periods for volume-related indicators

time_periods_volume = [5, 10, 14, 20, 30, 50, 100]

# Money Flow Index (MFI)

for period in time_periods_volume:
data[f'mfi_{period}'] = ta.MFI(data['high'], data['low'], data['close'], data['

# Volume Rate of Change (VROC)

for period in time_periods_volume:
data[f'vroc_{period}'] = ta.ROC(data['volume'], timeperiod=period)

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 32/132

8/22/24, 11:05 PM Project-ML-CODE

C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4136467864.py:6: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'mfi_{period}'] = ta.MFI(data['high'], data['low'], data['close'], data['vol
ume'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4136467864.py:6: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'mfi_{period}'] = ta.MFI(data['high'], data['low'], data['close'], data['vol
ume'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4136467864.py:6: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'mfi_{period}'] = ta.MFI(data['high'], data['low'], data['close'], data['vol
ume'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4136467864.py:6: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'mfi_{period}'] = ta.MFI(data['high'], data['low'], data['close'], data['vol
ume'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4136467864.py:6: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'mfi_{period}'] = ta.MFI(data['high'], data['low'], data['close'], data['vol
ume'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4136467864.py:6: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'mfi_{period}'] = ta.MFI(data['high'], data['low'], data['close'], data['vol
ume'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4136467864.py:6: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data[f'mfi_{period}'] = ta.MFI(data['high'], data['low'], data['close'], data['vol
ume'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4136467864.py:10: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'vroc_{period}'] = ta.ROC(data['volume'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4136467864.py:10: PerformanceWarni

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 33/132

8/22/24, 11:05 PM Project-ML-CODE

ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'vroc_{period}'] = ta.ROC(data['volume'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4136467864.py:10: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'vroc_{period}'] = ta.ROC(data['volume'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4136467864.py:10: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'vroc_{period}'] = ta.ROC(data['volume'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4136467864.py:10: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'vroc_{period}'] = ta.ROC(data['volume'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4136467864.py:10: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'vroc_{period}'] = ta.ROC(data['volume'], timeperiod=period)
C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\4136467864.py:10: PerformanceWarni
ng: DataFrame is highly fragmented. This is usually the result of calling `frame.in
sert` many times, which has poor performance. Consider joining all columns at once
using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = fram
e.copy()`
data[f'vroc_{period}'] = ta.ROC(data['volume'], timeperiod=period)

Step 5.1: Drop NaN values

Ensure there are no NaN values left after feature generation.

In [ ]: # Final check to ensure no NaN values are left after feature generation
data = data.dropna()

# Display the first few rows of the final data with all features
data.head()

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 34/132

8/22/24, 11:05 PM Project-ML-CODE

Out[ ]: date open high close low volume bb_upper_5 bb_m

2012-
2713 157.575019 158.216029 156.313030 156.146100 1626181.0 167.051730 160
11-08

2012-
2714 156.306352 157.448152 156.787110 156.019233 1167006.0 162.486557 158
11-09

2012-
2715 156.913977 157.434798 154.550251 152.834213 2046433.0 159.423325 156
11-12

2012-
2716 154.243101 154.243101 152.280007 150.704190 2044777.0 159.967128 155
11-13

2012-
2717 152.493677 154.243101 154.009399 151.238365 1223848.0 158.046785 154
11-14

5 rows × 295 columns

Step 5.2: Cut data to start from 2013

Filter the data to focus on the period from 2013 to 2019.

In [ ]: # Cut the data to start from 2013

start_date = '2013-01-01'
end_date = '2019-12-31'
data = data[(data['date'] >= start_date) & (data['date'] <= end_date)]

data.head()

Out[ ]: date open high close low volume bb_upper_5 bb_m

2013-
2751 141.556439 141.556439 138.177780 136.882405 5837034.0 142.200099 139
01-04

2013-
2752 137.469998 137.469998 136.408325 135.213108 4127139.0 142.415461 138
01-07

2013-
2753 135.667157 140.788562 140.781885 135.667157 6528698.0 143.124203 139
01-08

2013-
2754 141.222579 145.329051 144.527788 141.222579 5374932.0 145.366993 139
01-09

2013-
2755 143.946873 144.895034 143.225736 142.224158 3052786.0 146.676703 140
01-10

5 rows × 295 columns

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 35/132

8/22/24, 11:05 PM Project-ML-CODE

Step 6.1: Calculate the future return

Compute the future return as the percentage change from t0 to t1 (t+1/t0).

In [ ]: # Calculate the future return as the percentage change from t0 to t1 (t+1/t0)

data['FUTURE_RETURN'] = (data['close'].shift(-1) - data['close']) / data['close']

# Drop the last row as it will have NaN value for FUTURE_RETURN
data = data.dropna(subset=['FUTURE_RETURN'])

# Display the first few rows to check the FUTURE_RETURN column

data[['date', 'close', 'FUTURE_RETURN']].head()

C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\933941594.py:2: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data['FUTURE_RETURN'] = (data['close'].shift(-1) - data['close']) / data['close']
Out[ ]: date close FUTURE_RETURN

2751 2013-01-04 138.177780 -0.012806

2752 2013-01-07 136.408325 0.032062

2753 2013-01-08 140.781885 0.026608

2754 2013-01-09 144.527788 -0.009009

2755 2013-01-10 143.225736 -0.015105

Step 6.2: Determine the threshold for

classification
Set the threshold for classification based on the distribution of future returns.

In [ ]: # Plot the distribution of returns

plt.figure(figsize=(10, 6))
sns.histplot(data['FUTURE_RETURN'], bins=100, kde=True)
plt.title('Distribution of Future Returns')
plt.xlabel('FUTURE_RETURN')
plt.ylabel('Frequency')
plt.xlim(-0.05, 0.05)
plt.show()

# Statistical description of the returns

print(data['FUTURE_RETURN'].describe())

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 36/132

8/22/24, 11:05 PM Project-ML-CODE

threshold = data['FUTURE_RETURN'].mean() + data['FUTURE_RETURN'].std()*0.5

print(f"Threshold for positive future return classification: {threshold:.4f}")

c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):

count 1701.000000
mean 0.001422
std 0.020056
min -0.100000
25% -0.009021
50% 0.000215
75% 0.011151
max 0.100000
Name: FUTURE_RETURN, dtype: float64
Threshold for positive future return classification: 0.0114

In [ ]: # Explain the threshold calculation

display(Math(r'\text{Threshold} = \mu + 0.5 \times \sigma'))

Threshold = μ + 0.5 × σ

Step 6.3: Create the target variable

Create the binary target variable y based on the threshold for positive future return.

In [ ]: # Explain the threshold calculation

binary_target_tree_latex = r"""
\begin{array}{c}

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 37/132

8/22/24, 11:05 PM Project-ML-CODE

\text{Start with FUTURE\_RETURN} \\

\downarrow \\
\text{Is } \text{FUTURE\_RETURN} > \text{threshold?} \\
\begin{array}{ll}
\text{Yes} & \quad \text{No} \\
\downarrow & \quad \downarrow \\
\text{y = 1 (Positive Return)} & \quad \text{y = 0 (No Positive Return)} \\
\end{array}
\end{array}
"""

display(Math(binary_target_tree_latex))

Start with FUTURE\_RETURN

↓
Is FUTURE\_RETURN> threshold?
Yes No
↓ ↓
y = 1 (Positive Return) y = 0 (No Positive Return)

In [ ]: # Create the binary target variable y: 1 if return > threshold, 0 otherwise

data['y'] = np.where(data['FUTURE_RETURN'] > threshold, 1, 0)

# Display the first few rows to verify the target variable

data[['date', 'close', 'FUTURE_RETURN', 'y']].head()

C:\Users\bob19\AppData\Local\Temp\ipykernel_21596\326017906.py:2: PerformanceWarnin
g: DataFrame is highly fragmented. This is usually the result of calling `frame.ins
ert` many times, which has poor performance. Consider joining all columns at once u
sing pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.
copy()`
data['y'] = np.where(data['FUTURE_RETURN'] > threshold, 1, 0)
Out[ ]: date close FUTURE_RETURN y

2751 2013-01-04 138.177780 -0.012806 0

2752 2013-01-07 136.408325 0.032062 1

2753 2013-01-08 140.781885 0.026608 1

2754 2013-01-09 144.527788 -0.009009 0

2755 2013-01-10 143.225736 -0.015105 0

In [ ]: # Check for class imbalance

data['y'].value_counts(normalize=True)

Out[ ]: y
0 0.753674
1 0.246326
Name: proportion, dtype: float64

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 38/132

8/22/24, 11:05 PM Project-ML-CODE

Step 7.1: Preprocessing - Scaling and

transforming features
Use pipelines to preprocess and scale the features before feeding them into the model.

In [ ]: # Explain StandardScaler
display(Math(r'\text{StandardScaler: } z = \frac{x - \mu}{\sigma}'))

# Explain MinMaxScaler
display(Math(r'\text{MinMaxScaler: } x_{scaled} = \frac{x - x_{min}}{x_{max} - x_{m

x−μ
StandardScaler: z =
σ
x − xmin
MinMaxScaler: xscaled =
xmax − xmin

In [ ]: # 1. Continuous/Numerical features (price-based indicators like moving averages)

numerical_features = [col for col in data.columns if 'sma' in col or 'ema' in col o

# 2. Volatility-based features (ATR, Bollinger Bands)

volatility_features = [col for col in data.columns if 'atr' in col or 'bb' in col o

# 3. Volume-based features (OBV, MFI, VROC)

volume_features = [col for col in data.columns if 'obv' in col or 'mfi' in col or '

# 4. Momentum-based features (RSI, MOM, ROC)

momentum_features = [col for col in data.columns if 'rsi' in col or 'mom' in col or

# 5. Ratios (H2L, O2C)

ratio_features = ['h2l', 'o2c']

# 6. Price and volume features

price_volume_features = ['open', 'close', 'high', 'low', 'volume']

# 7. SMA slope and slope change features

sma_columns = ['sma_5', 'sma_10', 'sma_20', 'sma_60', 'sma_120', 'sma_200']
sma_slope_features = [f'{sma}_slope' for sma in sma_columns]
sma_slope_change_features = [f'{sma}_slope_change' for sma in sma_columns]

# Combine all feature lists

all_features = numerical_features + volatility_features + volume_features + momentu

# Pipeline for numerical features

numerical_pipeline = Pipeline(steps=[
('scaler', StandardScaler()) # Standard scaling for numerical features
])

# Pipeline for volatility features

volatility_pipeline = Pipeline(steps=[
('scaler', MinMaxScaler()) # Min-Max scaling for volatility-based features
])

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 39/132

8/22/24, 11:05 PM Project-ML-CODE

# Pipeline for volume-based features

volume_pipeline = Pipeline(steps=[
('scaler', MinMaxScaler()) # Min-Max scaling for volume-based features
])

# Pipeline for momentum-based features

momentum_pipeline = Pipeline(steps=[
('scaler', MinMaxScaler()) # Min-Max scaling for momentum features
])

# Pipeline for ratio features

ratio_pipeline = Pipeline(steps=[
('scaler', MinMaxScaler()) # Min-Max scaling for ratio features
])

# Pipeline for price and volume features

price_volume_pipeline = Pipeline(steps=[
('scaler', StandardScaler()) # Standard scaling for price and volume features
])

# Pipeline for SMA slope and slope change features

sma_slope_pipeline = Pipeline(steps=[
('scaler', StandardScaler()) # Standard scaling for SMA slope and slope change
])

# Combine all pipelines into a single ColumnTransformer

preprocessor = ColumnTransformer(
transformers=[
('num', numerical_pipeline, numerical_features),
('vol', volatility_pipeline, volatility_features),
('volm', volume_pipeline, volume_features),
('mom', momentum_pipeline, momentum_features),
('rat', ratio_pipeline, ratio_features),
('price_vol', price_volume_pipeline, price_volume_features), # Apply scali
('sma_slope', sma_slope_pipeline, sma_slope_features), # SMA slope feature
('sma_slope_change', sma_slope_pipeline, sma_slope_change_features) # SMA
]
)

# Apply the preprocessor to the data

data_processed = preprocessor.fit_transform(data)

# Convert the processed data back into a DataFrame

processed_columns = numerical_features + volatility_features + volume_features + mo
data_processed_df = pd.DataFrame(data_processed, columns=processed_columns)

# Add the target column 'y' back to the processed DataFrame

data_processed_df['y'] = data['y'].values

# drop nan values

data_processed_df = data_processed_df.dropna()

# Display the first few rows of the processed data

data_processed_df.head()

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 40/132

8/22/24, 11:05 PM Project-ML-CODE

Out[ ]: dema_5 dema_10 dema_14 dema_20 dema_30 dema_50 dema_60 ema_5

0 -0.815957 -0.814504 -0.815330 -0.818313 -0.820458 -0.813450 -0.807504 -0.813880 -0

1 -0.820938 -0.817867 -0.817796 -0.819957 -0.821721 -0.814863 -0.809009 -0.817156 -0

2 -0.815031 -0.815267 -0.815949 -0.818560 -0.820921 -0.815003 -0.809445 -0.814218 -0

3 -0.804461 -0.808981 -0.811168 -0.814946 -0.818530 -0.814112 -0.809004 -0.807873 -0

4 -0.801350 -0.805839 -0.808430 -0.812629 -0.816900 -0.813604 -0.808865 -0.805168 -0

5 rows × 173 columns

In [ ]: data_processed_df.describe()

Out[ ]: dema_5 dema_10 dema_14 dema_20 dema_30 de

count 1.701000e+03 1.701000e+03 1.701000e+03 1.701000e+03 1.701000e+03 1.70100

mean -6.683530e-17 1.336706e-16 2.005059e-16 -2.673412e-16 1.336706e-16 2.6734

std 1.000294e+00 1.000294e+00 1.000294e+00 1.000294e+00 1.000294e+00 1.00029

min -1.015105e+00 -1.015205e+00 -1.015182e+00 -1.013721e+00 -1.010513e+00 -1.00597

25% -8.376433e-01 -8.353190e-01 -8.344057e-01 -8.333103e-01 -8.327865e-01 -8.3885

50% -3.944266e-01 -3.966726e-01 -3.928047e-01 -3.887822e-01 -3.936933e-01 -4.1118

75% 8.155541e-01 8.166955e-01 8.125680e-01 8.145304e-01 8.188911e-01 8.0405

max 2.723433e+00 2.723200e+00 2.717185e+00 2.708961e+00 2.706439e+00 2.71539

8 rows × 173 columns

In [ ]: # Check for missing columns (features) in the processed data

missing_features = set(all_features) - set(data_processed_df.columns)
if missing_features:
print(f"Missing features that were not scaled: {missing_features}")
else:
print("All features have been scaled and included in the processed data.")

All features have been scaled and included in the processed data.

Exploratory Data Analysis (EDA)

In [ ]: # Ensure 'date' column is in datetime format
data['date'] = pd.to_datetime(data['date'])

# Plotting price and volume trends over time

def plot_price_volume_trends(data):

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 41/132

8/22/24, 11:05 PM Project-ML-CODE

plt.figure(figsize=(14, 8))

# Subplot 1: Closing Price

plt.subplot(2, 1, 1)
plt.plot(data['date'], data['close'], label='Closing Price', color='blue')
plt.title('Closing Price Over Time')
plt.xlabel('Date')
plt.ylabel('Price')
plt.grid(True)

# Subplot 2: Volume
plt.subplot(2, 1, 2)
plt.plot(data['date'], data['volume'], label='Volume', color='orange')
plt.title('Volume Over Time')
plt.xlabel('Date')
plt.ylabel('Volume')
plt.grid(True)

plt.tight_layout()
plt.show()

plot_price_volume_trends(data)

In [ ]: # Define a function to plot the distribution

def plot_pairwise_relationships_numerical(data, features):
sns.pairplot(data[features], diag_kind='kde', plot_kws={'alpha':0.5})
plt.suptitle('Pairplot of Numerical Features', y=1.02)
plt.show()

# Select a subset of numerical features for pairplot

pairplot_numerical_features = numerical_features[:5]

plot_pairwise_relationships_numerical(data, pairplot_numerical_features)

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 42/132

8/22/24, 11:05 PM Project-ML-CODE

c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 43/132

8/22/24, 11:05 PM Project-ML-CODE

In [ ]: def plot_pairwise_relationships_volatility(data, features):

sns.pairplot(data[features], diag_kind='kde', plot_kws={'alpha':0.5})
plt.suptitle('Pairplot of Volatility Features', y=1.02)
plt.show()

# Select a subset of volatility features for pairplot

pairplot_volatility_features = volatility_features[:5]

plot_pairwise_relationships_volatility(data, pairplot_volatility_features)

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 44/132

8/22/24, 11:05 PM Project-ML-CODE

c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 45/132

8/22/24, 11:05 PM Project-ML-CODE

In [ ]: def plot_pairwise_relationships_volume(data, features):

sns.pairplot(data[features], diag_kind='kde', plot_kws={'alpha':0.5})
plt.suptitle('Pairplot of Volume Features', y=1.02)
plt.show()

# Select a subset of volume features for pairplot

pairplot_volume_features = volume_features[:5]

plot_pairwise_relationships_volume(data, pairplot_volume_features)

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 46/132

8/22/24, 11:05 PM Project-ML-CODE

c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 47/132

8/22/24, 11:05 PM Project-ML-CODE

In [ ]: def plot_pairwise_relationships_momentum(data, features):

sns.pairplot(data[features], diag_kind='kde', plot_kws={'alpha':0.5})
plt.suptitle('Pairplot of Momentum Features', y=1.02)
plt.show()

# Select a subset of momentum features for pairplot

pairplot_momentum_features = momentum_features[:5]

plot_pairwise_relationships_momentum(data, pairplot_momentum_features)

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 48/132

8/22/24, 11:05 PM Project-ML-CODE

c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 49/132

8/22/24, 11:05 PM Project-ML-CODE

In [ ]: def plot_pairwise_relationships_ratio(data, features):

sns.pairplot(data[features], diag_kind='kde', plot_kws={'alpha':0.5})
plt.suptitle('Pairplot of Ratio Features', y=1.02)
plt.show()

# Select a subset of ratio features for pairplot

pairplot_ratio_features = ratio_features

plot_pairwise_relationships_ratio(data, pairplot_ratio_features)

c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 50/132

8/22/24, 11:05 PM Project-ML-CODE

In [ ]: # Step 1: Split the data into success and failure groups

data_success = data[data['y'] == 1] # success
data_failure = data[data['y'] == 0] # failure

# Step 2: Plot the distribution of the selected features

def plot_simple_distribution(data1, data2, features):
num_features = len(features)
num_rows = (num_features // 2) + 1 if num_features % 2 != 0 else num_features /

plt.figure(figsize=(14, num_rows * 6))

for i, feature in enumerate(features):

plt.subplot(num_rows, 2, i + 1)
sns.histplot(data1[feature], stat="density", color='red', kde=True, label='
sns.histplot(data2[feature], stat="density", color='blue', kde=True, label=

plt.title(f'Distribution of {feature}')
plt.legend()

plt.tight_layout()
plt.show()

plot_simple_distribution(data_success, data_failure, numerical_features)

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 51/132

8/22/24, 11:05 PM Project-ML-CODE

c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 52/132

8/22/24, 11:05 PM Project-ML-CODE

c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 53/132

8/22/24, 11:05 PM Project-ML-CODE

c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 54/132

8/22/24, 11:05 PM Project-ML-CODE

c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 55/132

8/22/24, 11:05 PM Project-ML-CODE

c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 56/132

8/22/24, 11:05 PM Project-ML-CODE

c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 57/132

8/22/24, 11:05 PM Project-ML-CODE

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 58/132

8/22/24, 11:05 PM Project-ML-CODE

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 59/132

8/22/24, 11:05 PM Project-ML-CODE

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 60/132

8/22/24, 11:05 PM Project-ML-CODE

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 61/132

8/22/24, 11:05 PM Project-ML-CODE

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 62/132

8/22/24, 11:05 PM Project-ML-CODE

In [ ]: def plot_volatility_features_distribution(data1, data2, features):

num_features = len(features)
num_rows = (num_features // 2) + 1 if num_features % 2 != 0 else num_features /

plt.figure(figsize=(14, num_rows * 6))

for i, feature in enumerate(features):

plt.subplot(num_rows, 2, i + 1)
sns.histplot(data1[feature], stat="density", color='red', kde=True, label='
sns.histplot(data2[feature], stat="density", color='blue', kde=True, label=

plt.title(f'Distribution of {feature}')
plt.legend()

plt.tight_layout()
plt.show()

plot_volatility_features_distribution(data_success, data_failure, volatility_featur

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 63/132

8/22/24, 11:05 PM Project-ML-CODE

c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 64/132

8/22/24, 11:05 PM Project-ML-CODE

c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 65/132

8/22/24, 11:05 PM Project-ML-CODE

c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 66/132

8/22/24, 11:05 PM Project-ML-CODE

c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 67/132

8/22/24, 11:05 PM Project-ML-CODE

c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 68/132

8/22/24, 11:05 PM Project-ML-CODE

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 69/132

8/22/24, 11:05 PM Project-ML-CODE

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 70/132

8/22/24, 11:05 PM Project-ML-CODE

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 71/132

8/22/24, 11:05 PM Project-ML-CODE

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 72/132

8/22/24, 11:05 PM Project-ML-CODE

In [ ]: def plot_volume_features_distribution(data1, data2, features):

num_features = len(features)

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 73/132

8/22/24, 11:05 PM Project-ML-CODE

num_rows = (num_features // 2) + 1 if num_features % 2 != 0 else num_features /

plt.figure(figsize=(14, num_rows * 6))

for i, feature in enumerate(features):

plt.subplot(num_rows, 2, i + 1)
sns.histplot(data1[feature], stat="density", color='red', kde=True, label='
sns.histplot(data2[feature], stat="density", color='blue', kde=True, label=

plt.title(f'Distribution of {feature}')
plt.legend()

plt.tight_layout()
plt.show()

plot_volume_features_distribution(data_success, data_failure, volume_features)

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 74/132

8/22/24, 11:05 PM Project-ML-CODE

c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 75/132

8/22/24, 11:05 PM Project-ML-CODE

c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 76/132

8/22/24, 11:05 PM Project-ML-CODE

c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 77/132

8/22/24, 11:05 PM Project-ML-CODE

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 78/132

8/22/24, 11:05 PM Project-ML-CODE

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 79/132

8/22/24, 11:05 PM Project-ML-CODE

In [ ]: def plot_momentum_features_distribution(data1, data2, features):

num_features = len(features)
num_rows = (num_features // 2) + 1 if num_features % 2 != 0 else num_features /

plt.figure(figsize=(14, num_rows * 6))

for i, feature in enumerate(features):

plt.subplot(num_rows, 2, i + 1)
sns.histplot(data1[feature], stat="density", color='red', kde=True, label='
sns.histplot(data2[feature], stat="density", color='blue', kde=True, label=

plt.title(f'Distribution of {feature}')
plt.legend()

plt.tight_layout()
plt.show()

plot_momentum_features_distribution(data_success, data_failure, momentum_features)

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 80/132

8/22/24, 11:05 PM Project-ML-CODE

c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 81/132

8/22/24, 11:05 PM Project-ML-CODE

c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 82/132

8/22/24, 11:05 PM Project-ML-CODE

c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 83/132

8/22/24, 11:05 PM Project-ML-CODE

c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 84/132

8/22/24, 11:05 PM Project-ML-CODE

c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 85/132

8/22/24, 11:05 PM Project-ML-CODE

c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 86/132

8/22/24, 11:05 PM Project-ML-CODE

c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 87/132

8/22/24, 11:05 PM Project-ML-CODE

c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 88/132

8/22/24, 11:05 PM Project-ML-CODE

c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 89/132

8/22/24, 11:05 PM Project-ML-CODE

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 90/132

8/22/24, 11:05 PM Project-ML-CODE

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 91/132

8/22/24, 11:05 PM Project-ML-CODE

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 92/132

8/22/24, 11:05 PM Project-ML-CODE

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 93/132

8/22/24, 11:05 PM Project-ML-CODE

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 94/132

8/22/24, 11:05 PM Project-ML-CODE

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 95/132

8/22/24, 11:05 PM Project-ML-CODE

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 96/132

8/22/24, 11:05 PM Project-ML-CODE

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 97/132

8/22/24, 11:05 PM Project-ML-CODE

In [ ]: def plot_ratio_features_distribution(data1, data2, features):

num_features = len(features)
num_rows = (num_features // 2) + 1 if num_features % 2 != 0 else num_features /

plt.figure(figsize=(14, num_rows * 6))

for i, feature in enumerate(features):

plt.subplot(num_rows, 2, i + 1)
sns.histplot(data1[feature], stat="density", color='red', kde=True, label='
sns.histplot(data2[feature], stat="density", color='blue', kde=True, label=

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 98/132

8/22/24, 11:05 PM Project-ML-CODE

plt.title(f'Distribution of {feature}')
plt.legend()

plt.tight_layout()
plt.show()

plot_ratio_features_distribution(data_success, data_failure, ratio_features)

c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):

In [ ]: def plot_ratio_features_distribution(data1, data2, features):

num_features = len(features)
num_rows = (num_features // 2) + 1 if num_features % 2 != 0 else num_features /

plt.figure(figsize=(14, num_rows * 6))

for i, feature in enumerate(features):

plt.subplot(num_rows, 2, i + 1)
sns.histplot(data1[feature], stat="density", color='red', kde=True, label='
sns.histplot(data2[feature], stat="density", color='blue', kde=True, label=

plt.title(f'Distribution of {feature}')
plt.legend()

plt.tight_layout()
plt.show()

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 99/132

8/22/24, 11:05 PM Project-ML-CODE

plot_ratio_features_distribution(data_success, data_failure, sma_slope_features)

c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 100/132

8/22/24, 11:05 PM Project-ML-CODE

In [ ]: def plot_ratio_features_distribution(data1, data2, features):

num_features = len(features)
num_rows = (num_features // 2) + 1 if num_features % 2 != 0 else num_features /

plt.figure(figsize=(14, num_rows * 6))

for i, feature in enumerate(features):

plt.subplot(num_rows, 2, i + 1)
sns.histplot(data1[feature], stat="density", color='red', kde=True, label='

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 101/132

8/22/24, 11:05 PM Project-ML-CODE

sns.histplot(data2[feature], stat="density", color='blue', kde=True, label=

plt.title(f'Distribution of {feature}')
plt.legend()

plt.tight_layout()
plt.show()

plot_ratio_features_distribution(data_success, data_failure, sma_slope_change_featu

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 102/132

8/22/24, 11:05 PM Project-ML-CODE

c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):
c:\Users\bob19\anaconda3\Lib\site-packages\seaborn\_oldcore.py:1119: FutureWarning:
use_inf_as_na option is deprecated and will be removed in a future version. Convert
inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 103/132

8/22/24, 11:05 PM Project-ML-CODE

In [ ]: # Correlation heatmap for numerical features

def plot_group_correlation_heatmap(data, features, title):
correlation_matrix = data[features].corr()
plt.figure(figsize=(10, 8))
sns.heatmap(correlation_matrix, annot=False, cmap='coolwarm', fmt='.2f', linewi
plt.title(f'{title} Correlation Heatmap')
plt.show()

# Call the function for each feature group

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 104/132

8/22/24, 11:05 PM Project-ML-CODE

plot_group_correlation_heatmap(data, numerical_features, 'Numerical Features')

plot_group_correlation_heatmap(data, volatility_features, 'Volatility Features')
plot_group_correlation_heatmap(data, volume_features, 'Volume Features')
plot_group_correlation_heatmap(data, momentum_features, 'Momentum Features')
plot_group_correlation_heatmap(data, ratio_features, 'Ratio Features')
plot_group_correlation_heatmap(data, sma_slope_features, 'SMA Slope Features')
plot_group_correlation_heatmap(data, sma_slope_change_features, 'SMA Slope Change F

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 105/132

8/22/24, 11:05 PM Project-ML-CODE

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 106/132

8/22/24, 11:05 PM Project-ML-CODE

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 107/132

8/22/24, 11:05 PM Project-ML-CODE

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 108/132

8/22/24, 11:05 PM Project-ML-CODE

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 109/132

8/22/24, 11:05 PM Project-ML-CODE

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 110/132

8/22/24, 11:05 PM Project-ML-CODE

In [ ]: # Calculate the correlation of all features with the target variable `y`
correlation_with_y = data.corr()['y'].sort_values(ascending=False)

# Select top 30 features most correlated with the target variable

top_features = correlation_with_y.index[1:31]

# Correlation heatmap for top features

def plot_top_correlation_heatmap(data, top_features):
correlation_matrix = data[top_features].corr()
plt.figure(figsize=(12, 10))
sns.heatmap(correlation_matrix, annot=False, cmap='coolwarm', fmt='.2f', linewi
plt.title('Top Correlated Features with y')
plt.show()

plot_top_correlation_heatmap(data, top_features)

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 111/132

8/22/24, 11:05 PM Project-ML-CODE

Feature Selection
1. Random Forest Feature Selection
In [ ]: # Explain feature importance
display(Math(r'\text{Feature Importance: } I(f_j) = \frac{1}{T} \sum_{t=1}^{T} I_t(

display(Markdown("""
**Explanation:**

- **Feature Importance**: Denoted as \( I(f_j) \), it represents the importance of

"""))

1 T
Feature Importance: I(fj ) = ∑ I (f )
T t=1 t j

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 112/132

8/22/24, 11:05 PM Project-ML-CODE

Explanation:

Feature Importance: Denoted as ( I(f_j) ), it represents the importance of feature ( f_j ) in

the model.

In [ ]: X = data_processed_df.drop(columns=['y'])
y = data_processed_df['y']

model = RandomForestClassifier(n_estimators=100, random_state=88)

model.fit(X, y)

importance_scores = model.feature_importances_
feature_importance_df = pd.DataFrame({'Feature': X.columns, 'Importance': importanc
feature_importance_df = feature_importance_df.sort_values(by='Importance', ascendin

# Select top 35% most important features

top_80_percent_features = feature_importance_df['Feature'].head(int(len(importance_
data_top_80_percent_df = X[top_80_percent_features]

print(f"Number of features after Random Forest selection: {len(top_80_percent_featu

Number of features after Random Forest selection: 60

2. SOM Feature Selection

In [ ]: # Explain SOM weight update
display(Math(r'\Delta w_{ij} = \eta(t) \times h_{ij}(t) \times (x(t) - w_{ij}(t))')

display(Markdown("""
**Explanation:**

- **SOM Weight Update**: This formula describes how the weights \( w_{ij} \) of the

Δwij = η(t) × hij (t) × (x(t) − wij (t))

Explanation:

SOM Weight Update: This formula describes how the weights ( w_{ij} ) of the nodes in a
Self-Organizing Map (SOM) are updated during training.

In [ ]: # Step 6.2: Retain 50-60% of features after SOM reduction

som = MiniSom(x=15, y=15, input_len=data_top_80_percent_df.shape[1], sigma=1.0, lea
som.random_weights_init(data_top_80_percent_df.values)
som.train_random(data_top_80_percent_df.values, 100)

# Compute feature importance based on SOM weights variance and select top 80%
som_feature_importance = np.var(som.get_weights(), axis=0).sum(axis=1)
sorted_som_indices = np.argsort(som_feature_importance)[::-1]
top_50_60_percent_indices = sorted_som_indices[:int(len(sorted_som_indices) * 0.8)]
top_som_features = data_top_80_percent_df.columns[top_50_60_percent_indices]

data_top_som_features_df = data_top_80_percent_df[top_som_features]

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 113/132

8/22/24, 11:05 PM Project-ML-CODE

print(f"Number of features after SOM reduction: {len(top_som_features)}")

Number of features after SOM reduction: 12

3. Leave-One-Feature-Out Feature Evaluation

In [ ]: feature_importance_scores = []
for feature in data_top_som_features_df.columns:
X_temp = data_top_som_features_df.drop(columns=[feature])
scores = cross_val_score(model, X_temp, y, cv=5, scoring='accuracy')
feature_importance_scores.append((feature, np.mean(scores)))

# Sort features by their importance scores

sorted_scores = sorted(feature_importance_scores, key=lambda x: x[1], reverse=True)

# Ensure at least 20 features are selected and avoid duplicates

final_features = []
for feature, score in sorted_scores:
if feature not in final_features:
final_features.append(feature)
if len(final_features) >= 20: # Stop when 20 unique features are selected
break

print(f"Number of features after Leave-One-Feature-Out evaluation: {len(final_featu

# display the selected features

print("Selected features after leave-one-feature-out evaluation:")
for feature in final_features:
print(feature)

Number of features after Leave-One-Feature-Out evaluation: 8

Selected features after leave-one-feature-out evaluation:
vroc_50
sma_5_slope
volume
mfi_14
return
natr_5
vroc_30
h2l

4. Mannually add back important yet uncaptured features

In [ ]: additional_features = [
'sma_5_slope',
'sma_5_slope_change',
'sma_10_slope',
'sma_10_slope_change',
'rsi_14',
'sma_20_slope',
'sma_20_slope_change'
]

# make sure the additional features are in the data

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 114/132

8/22/24, 11:05 PM Project-ML-CODE

for feature in additional_features:

if feature not in final_features and feature in data_processed_df.columns:
final_features.append(feature)
if len(final_features) >= 15: # stop when 15 features are selected
break

print(f"Number of features after manual addition: {len(final_features)}")

# display the final selected features

print("Final selected features:")
for feature in final_features:
print(feature)

Number of features after manual addition: 14

Final selected features:
vroc_50
sma_5_slope
volume
mfi_14
return
natr_5
vroc_30
h2l
sma_5_slope_change
sma_10_slope
sma_10_slope_change
rsi_14
sma_20_slope
sma_20_slope_change

Part Five: Model Building and Optimization

Step 7.2: Split the data into training,

validation, and test sets
Sequentially split the data to maintain time order, avoiding data leakage.

In [ ]: # Define the features and target

X = data_processed_df[final_features] # Selected features
y = data_processed_df['y'] # Target variable

# Split into train (60%) and temp (40%)

X_train, X_temp, y_train, y_temp = train_test_split(X, y, train_size=0.6, shuffle=F

# Split temp into validation (20%) and test (20%)

X_val, X_test, y_val, y_test = train_test_split(X_temp, y_temp, train_size=0.5, shu

train_shape = X_train.shape
val_shape = X_val.shape
test_shape = X_test.shape

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 115/132

8/22/24, 11:05 PM Project-ML-CODE

# Create a dictionary with the shapes of the datasets

shape_dict = {
"Dataset": ["Training Set", "Validation Set", "Test Set"],
"Shape": [train_shape, val_shape, test_shape]
}

# Create a DataFrame from the dictionary

shape_df = pd.DataFrame(shape_dict)

shape_df

Out[ ]: Dataset Shape

0 Training Set (1020, 22)

1 Validation Set (340, 22)

2 Test Set (341, 22)

In [ ]: # Display the first few rows of each set to ensure data integrity
print("First few rows of the training set:")
print(X_train.head())
print(y_train.head())

print('----------------------------------')

print("First few rows of the validation set:")

print(X_val.head())
print(y_val.head())

print('----------------------------------')

print("First few rows of the test set:")

print(X_test.head())
print(y_test.head())

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 116/132

8/22/24, 11:05 PM Project-ML-CODE

First few rows of the training set:

vroc_50 vroc_50 sma_5_slope sma_5_slope volume mfi_14 return \
0 0.238114 0.238114 -0.364965 -0.364965 0.696042 0.446488 -0.566936
1 0.287847 0.287847 -0.252934 -0.252934 0.018891 0.462097 -0.709286
2 0.549192 0.549192 -0.133794 -0.133794 0.969954 0.575913 1.528269
3 0.165723 0.165723 -0.062294 -0.062294 0.513041 0.564499 1.256258
4 0.159367 0.159367 -0.053494 -0.053494 -0.406574 0.499435 -0.519949

natr_5 vroc_30 vroc_30 ... sma_5_slope_change sma_10_slope \

0 0.299184 0.151988 0.151988 ... 0.286730 -0.278774
1 0.269562 0.121151 0.121151 ... 0.315569 -0.325432
2 0.266935 0.124195 0.124195 ... 0.335590 -0.376946
3 0.254584 0.180026 0.180026 ... 0.201405 -0.401666
4 0.224688 0.137091 0.137091 ... 0.024797 -0.336236

sma_10_slope sma_10_slope_change sma_10_slope_change rsi_14 \

0 -0.278774 -0.255187 -0.255187 0.397164
1 -0.325432 -0.218168 -0.218168 0.363216
2 -0.376946 -0.240890 -0.240890 0.465412
3 -0.401666 -0.115535 -0.115535 0.538914
4 -0.336236 0.306232 0.306232 0.510117

sma_20_slope sma_20_slope sma_20_slope_change sma_20_slope_change

0 -0.329610 -0.329610 0.598738 0.598738
1 -0.254989 -0.254989 0.497388 0.497388
2 -0.153306 -0.153306 0.677619 0.677619
3 -0.076604 -0.076604 0.511252 0.511252
4 -0.051910 -0.051910 0.164891 0.164891

[5 rows x 22 columns]
0 0
1 1
2 1
3 0
4 0
Name: y, dtype: int32
----------------------------------
First few rows of the validation set:
vroc_50 vroc_50 sma_5_slope sma_5_slope volume mfi_14 \
1020 0.066409 0.066409 0.256744 0.256744 -0.366256 0.880916
1021 0.177007 0.177007 0.507654 0.507654 0.303612 0.898204
1022 0.049973 0.049973 0.647739 0.647739 0.187647 0.834521
1023 0.060037 0.060037 0.762405 0.762405 -0.249079 0.749484
1024 0.033754 0.033754 0.759129 0.759129 -0.297278 0.667080

return natr_5 vroc_30 vroc_30 ... sma_5_slope_change \

1020 0.974213 0.089731 0.202074 0.202074 ... 0.072243
1021 0.910180 0.103433 0.238673 0.238673 ... 0.706748
1022 -0.011183 0.114553 0.241149 0.241149 ... 0.394587
1023 -0.663577 0.116055 0.146385 0.146385 ... 0.322990
1024 -0.450654 0.118912 0.136383 0.136383 ... -0.009218

sma_10_slope sma_10_slope sma_10_slope_change sma_10_slope_change \

1020 0.506244 0.506244 0.446649 0.446649
1021 0.593454 0.593454 0.408131 0.408131
1022 0.659972 0.659972 0.311326 0.311326

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 117/132

8/22/24, 11:05 PM Project-ML-CODE

1023 0.674107 0.674107 0.066246 0.066246

1024 0.637897 0.637897 -0.169289 -0.169289

rsi_14 sma_20_slope sma_20_slope sma_20_slope_change \

1020 0.882277 0.305083 0.305083 0.106389
1021 0.928890 0.381660 0.381660 0.510416
1022 0.931446 0.412521 0.412521 0.205960
1023 0.809778 0.419277 0.419277 0.045434
1024 0.740583 0.401451 0.401451 -0.118283

sma_20_slope_change
1020 0.106389
1021 0.510416
1022 0.205960
1023 0.045434
1024 -0.118283

[5 rows x 22 columns]
1020 1
1021 0
1022 0
1023 0
1024 0
Name: y, dtype: int32
----------------------------------
First few rows of the test set:
vroc_50 vroc_50 sma_5_slope sma_5_slope volume mfi_14 \
1360 0.066694 0.066694 -2.707663 -2.707663 -0.067671 0.154347
1361 0.052262 0.052262 -2.822225 -2.822225 -0.439935 0.221686
1362 0.067878 0.067878 -2.405615 -2.405615 -0.249834 0.305615
1363 0.021536 0.021536 -1.617521 -1.617521 -0.672310 0.378990
1364 0.069992 0.069992 -0.930498 -0.930498 -0.246481 0.273434

return natr_5 vroc_30 vroc_30 ... sma_5_slope_change \

1360 1.165500 0.233288 0.079422 0.079422 ... -0.694504
1361 -0.801700 0.225490 0.045841 0.045841 ... -0.322677
1362 1.084222 0.232434 0.062609 0.062609 ... 1.173477
1363 -0.409441 0.198187 0.067305 0.067305 ... 2.219834
1364 -0.648342 0.207961 0.087926 0.087926 ... 1.935149

sma_10_slope sma_10_slope sma_10_slope_change sma_10_slope_change \

1360 -2.108801 -2.108801 -1.513420 -1.513420
1361 -2.447012 -2.447012 -1.582201 -1.582201
1362 -2.436736 -2.436736 0.048196 0.048196
1363 -2.356027 -2.356027 0.377714 0.377714
1364 -2.188009 -2.188009 0.786191 0.786191

rsi_14 sma_20_slope sma_20_slope sma_20_slope_change \

1360 0.309545 -0.860509 -0.860509 -2.332326
1361 0.275045 -1.192977 -1.192977 -2.213717
1362 0.362596 -1.385926 -1.385926 -1.284555
1363 0.344629 -1.507684 -1.507684 -0.810445
1364 0.314571 -1.595994 -1.595994 -0.587683

sma_20_slope_change
1360 -2.332326

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 118/132

8/22/24, 11:05 PM Project-ML-CODE

1361 -2.213717
1362 -1.284555
1363 -0.810445
1364 -0.587683

[5 rows x 22 columns]
1360 0
1361 1
1362 0
1363 0
1364 0
Name: y, dtype: int32

Step 8.1: Hyperparameter tuning with

GridSearchCV
Perform hyperparameter tuning on Logistic Regression, Random Forest, and SVM models
using GridSearchCV.

In [ ]: # Explain the logistic regression function

display(Math(r'\text{Logistic Regression: } P(y=1|x) = \frac{1}{1 + e^{-(\beta_0 +

display(Markdown("""
**Explanation:**

- **Logistic Regression Function**: This formula represents the probability that th

"""))

1
Logistic Regression: P (y = 1|x) = −(β0 +β1 x1 +⋯+βn xn )
1+e
Explanation:

Logistic Regression Function: This formula represents the probability that the dependent
variable ( y ) equals 1, given the input features ( x ).

In [ ]: # Explain the SVM decision function

display(Math(r'\text{SVM Decision Function: } f(x) = w \cdot x + b'))

# Simple explanation in markdown

display(Markdown("""
**Explanation:**

- **SVM Decision Function**: This formula represents the decision boundary used by
"""))

SVM Decision Function: f(x) = w ⋅ x + b

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 119/132

8/22/24, 11:05 PM Project-ML-CODE

Explanation:

SVM Decision Function: This formula represents the decision boundary used by a Support
Vector Machine (SVM) to classify data points.

In [ ]: # Define parameter grids for each model

log_reg_params = {
'C': [1, 10, 20, 30],
'penalty': ['l2'],
'solver': ['liblinear']
}

rf_params = {
'n_estimators': [100, 200, 500],
'max_depth': [None, 10, 20],
'min_samples_split': [2, 5, 10],
'min_samples_leaf': [1, 2, 4]
}

svm_params = {
'C': [1, 10, 100],
'kernel': ['linear', 'rbf'],
'gamma': ['scale']
}

# Step 8.1: Perform hyperparameter tuning with GridSearchCV

log_reg_grid = GridSearchCV(LogisticRegression(random_state=88, class_weight='balan
rf_grid = GridSearchCV(RandomForestClassifier(random_state=88, class_weight='balanc
svm_grid = GridSearchCV(SVC(probability=True, random_state=88, class_weight='balanc

# Train with the best parameters found

log_reg_grid.fit(X_train, y_train)
rf_grid.fit(X_train, y_train)
svm_grid.fit(X_train, y_train)

# Get the best estimators

log_reg_best = log_reg_grid.best_estimator_
rf_best = rf_grid.best_estimator_
svm_best = svm_grid.best_estimator_

# Print best parameters

print("Best Logistic Regression Params:", log_reg_grid.best_params_)
print("Best Random Forest Params:", rf_grid.best_params_)
print("Best SVM Params:", svm_grid.best_params_)

Best Logistic Regression Params: {'C': 30, 'penalty': 'l2', 'solver': 'liblinear'}
Best Random Forest Params: {'max_depth': None, 'min_samples_leaf': 1, 'min_samples_s
plit': 2, 'n_estimators': 500}
Best SVM Params: {'C': 100, 'gamma': 'scale', 'kernel': 'linear'}

Step 8.2: Evaluate the tuned base learners

on the validation set
file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 120/132
8/22/24, 11:05 PM Project-ML-CODE

Evaluate the performance of tuned models on the validation set using accuracy, F1 score,
and AUC-ROC.

In [ ]: # Logistic Regression
y_pred_log_reg = log_reg_best.predict(X_val)
y_pred_log_reg_proba = log_reg_best.predict_proba(X_val)[:, 1]
log_reg_accuracy = accuracy_score(y_val, y_pred_log_reg)
log_reg_f1 = f1_score(y_val, y_pred_log_reg)
log_reg_auc = roc_auc_score(y_val, y_pred_log_reg_proba)

# Random Forest
y_pred_rf = rf_best.predict(X_val)
y_pred_rf_proba = rf_best.predict_proba(X_val)[:, 1]
rf_accuracy = accuracy_score(y_val, y_pred_rf)
rf_f1 = f1_score(y_val, y_pred_rf)
rf_auc = roc_auc_score(y_val, y_pred_rf_proba)

# SVM
y_pred_svm = svm_best.predict(X_val)
y_pred_svm_proba = svm_best.predict_proba(X_val)[:, 1]
svm_accuracy = accuracy_score(y_val, y_pred_svm)
svm_f1 = f1_score(y_val, y_pred_svm)
svm_auc = roc_auc_score(y_val, y_pred_svm_proba)

metrics_dict = {
'Model': ['Logistic Regression', 'Random Forest', 'SVM'],
'Validation Set Accuracy': [log_reg_accuracy, rf_accuracy, svm_accuracy],
'Validation Set F1 Score': [log_reg_f1, rf_f1, svm_f1],
'Validation Set AUC-ROC': [log_reg_auc, rf_auc, svm_auc]
}

metrics_df = pd.DataFrame(metrics_dict)

# Display the results

metrics_df.set_index('Model').T

Out[ ]: Model Logistic Regression Random Forest SVM

Validation Set Accuracy 0.641176 0.726471 0.679412

Validation Set F1 Score 0.377551 0.000000 0.347305

Validation Set AUC-ROC 0.567059 0.526303 0.577916

In [ ]: # Explain the formula for Accuracy

display(Math(r'\text{Accuracy: } \frac{TP + TN}{TP + TN + FP + FN}'))

display(Markdown("""
**Explanation:**

- **Accuracy**: This metric represents the proportion of correctly classified insta

"""))

# Explain the formula for F1 Score

display(Math(r'\text{F1 Score: } 2 \times \frac{\text{Precision} \times \text{Recal

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 121/132

8/22/24, 11:05 PM Project-ML-CODE

display(Markdown("""
**Explanation:**

- **F1 Score**: The F1 Score is a metric that balances precision and recall, provid
"""))

# AUC
display(Math(r'\text{AUC: Area Under the ROC Curve}'))

display(Markdown("""
**Explanation:**

- **AUC (Area Under the ROC Curve)**: AUC measures the area under the Receiver Oper
"""))

TP + TN
Accuracy:
TP + TN + FP + FN
Explanation:

Accuracy: This metric represents the proportion of correctly classified instances (both true
positives and true negatives) out of the total number of instances.

Precision × Recall
F1 Score: 2 ×
Precision + Recall
Explanation:

F1 Score: The F1 Score is a metric that balances precision and recall, providing a single
measure that accounts for both false positives and false negatives.

AUC: Area Under the ROC Curve

Explanation:

AUC (Area Under the ROC Curve): AUC measures the area under the Receiver Operating
Characteristic (ROC) curve, which plots the true positive rate (sensitivity) against the false
positive rate (1-specificity).

Part Seven: Building the Blending Model

Step 9.1: Generate predictions from tuned

base learners on the validation set
Use tuned models to generate predictions and probabilities on the validation set.

In [ ]: # Predict probabilities from tuned base learners

log_reg_val_pred = log_reg_best.predict_proba(X_val)[:, 1]
rf_val_pred = rf_best.predict_proba(X_val)[:, 1]

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 122/132

8/22/24, 11:05 PM Project-ML-CODE

svm_val_pred = svm_best.predict_proba(X_val)[:, 1]

# Create a new dataset using these predictions

blending_X_val = np.column_stack((log_reg_val_pred, rf_val_pred, svm_val_pred))
blending_y_val = y_val # The true labels remain the same

print("Shape of blending validation dataset:", blending_X_val.shape)

Shape of blending validation dataset: (340, 3)

Step 9.2: Train the blending model

Use Logistic Regression as a meta-learner to combine predictions from base models.

In [ ]: # Logistic Regression as the meta-learner

blending_model = LogisticRegression(random_state=88, class_weight='balanced')

# Train the meta-learner on the blending validation dataset

blending_model.fit(blending_X_val, blending_y_val)

Out[ ]: ▾ LogisticRegression i ?

LogisticRegression(class_weight='balanced', random_state=88)

Step 9.3: Use the Trained Blending Model to Predict

on the Test Set and Evaluate Performance
In [ ]: # Predict probabilities from tuned base learners on the test set
log_reg_test_pred = log_reg_best.predict_proba(X_test)[:, 1]
rf_test_pred = rf_best.predict_proba(X_test)[:, 1]
svm_test_pred = svm_best.predict_proba(X_test)[:, 1]

# Create a new dataset for the test set using these predictions
blending_X_test = np.column_stack((log_reg_test_pred, rf_test_pred, svm_test_pred))
blending_y_test = y_test # The true labels remain the same

# Step 9.3: Predict and evaluate the blending model on the test set

# Predict using the blending model

blending_test_pred = blending_model.predict(blending_X_test)
blending_test_pred_proba = blending_model.predict_proba(blending_X_test)[:, 1]

# Calculate performance metrics

blending_accuracy = accuracy_score(blending_y_test, blending_test_pred)
blending_f1 = f1_score(blending_y_test, blending_test_pred)
blending_auc = roc_auc_score(blending_y_test, blending_test_pred_proba)

metrics_dict = {
'Blending Model - Test Set Accuracy': [blending_accuracy],
'Blending Model - Test Set F1 Score': [blending_f1],
'Blending Model - Test Set AUC-ROC': [blending_auc]

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 123/132

8/22/24, 11:05 PM Project-ML-CODE

metrics_df = pd.DataFrame(metrics_dict)

# Display the results

metrics_df

Out[ ]: Blending Model - Test Set Blending Model - Test Set Blending Model - Test Set
Accuracy F1 Score AUC-ROC

0 0.521994 0.365759 0.534573

Step 10.1: Evaluate the blending model on

the test set
In [ ]: # Generate predictions and probabilities from the blending model on the test set
blending_test_pred = blending_model.predict(blending_X_test)
blending_test_pred_proba = blending_model.predict_proba(blending_X_test)[:, 1]

# Calculate evaluation metrics

blending_accuracy = accuracy_score(blending_y_test, blending_test_pred)
blending_precision = precision_score(blending_y_test, blending_test_pred)
blending_recall = recall_score(blending_y_test, blending_test_pred)
blending_f1 = f1_score(blending_y_test, blending_test_pred)
blending_auc = roc_auc_score(blending_y_test, blending_test_pred_proba)

# Generate the classification report as a dictionary

report_dict = classification_report(blending_y_test, blending_test_pred, output_dic

# Convert the dictionary to a DataFrame

report_df = pd.DataFrame(report_dict).transpose()

# Display the DataFrame

report_df

Out[ ]: precision recall f1-score support

0 0.752874 0.521912 0.616471 251.000000

1 0.281437 0.522222 0.365759 90.000000

accuracy 0.521994 0.521994 0.521994 0.521994

macro avg 0.517155 0.522067 0.491115 341.000000

weighted avg 0.628448 0.521994 0.550300 341.000000

In [ ]: # Calculate evaluation metrics

blending_accuracy = accuracy_score(blending_y_test, blending_test_pred)
blending_f1 = f1_score(blending_y_test, blending_test_pred)
blending_auc = roc_auc_score(blending_y_test, blending_test_pred_proba)

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 124/132

8/22/24, 11:05 PM Project-ML-CODE

# Confusion Matrix
conf_matrix = confusion_matrix(blending_y_test, blending_test_pred)
ConfusionMatrixDisplay(conf_matrix).plot()
plt.title("Confusion Matrix of Blending Model")
plt.show()

In [ ]: # Plot the ROC curve

fpr, tpr, _ = roc_curve(blending_y_test, blending_test_pred_proba)
plt.figure(figsize=(10, 7))
plt.plot(fpr, tpr, color='b', label=f'AUC = {blending_auc:.2f}')
plt.plot([0, 1], [0, 1], color='r', linestyle='--')
plt.title('Receiver Operating Characteristic (ROC) Curve')
plt.xlabel('False Positive Rate')
plt.ylabel('True Positive Rate')
plt.legend(loc='lower right')
plt.show()

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 125/132

8/22/24, 11:05 PM Project-ML-CODE

Step 11.1: Create backtest DataFrame

Generate a backtest DataFrame containing necessary columns for strategy evaluation.

In [ ]: # Predict probabilities from tuned base learners on the test set

log_reg_test_pred = log_reg_best.predict_proba(X_test)[:, 1]
rf_test_pred = rf_best.predict_proba(X_test)[:, 1]
svm_test_pred = svm_best.predict_proba(X_test)[:, 1]

# Create a new dataset for the test set using these predictions
blending_X_test = np.column_stack((log_reg_test_pred, rf_test_pred, svm_test_pred))
blending_y_test = y_test # The true labels remain the same

# Predict using the blending model

blending_test_pred = blending_model.predict(blending_X_test)
blending_test_pred_proba = blending_model.predict_proba(blending_X_test)[:, 1]

# Calculate performance metrics

blending_accuracy = accuracy_score(blending_y_test, blending_test_pred)
blending_f1 = f1_score(blending_y_test, blending_test_pred)
blending_auc = roc_auc_score(blending_y_test, blending_test_pred_proba)

# Print the performance of the blending model

print("Blending Model - Accuracy: {:.4f}, F1 Score: {:.4f}, AUC: {:.4f}".format(ble

Blending Model - Accuracy: 0.5220, F1 Score: 0.3658, AUC: 0.5346

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 126/132

8/22/24, 11:05 PM Project-ML-CODE

Step 11.2: Calculate daily strategy returns

Compute daily returns based on the shifted signal and the future returns.

In [ ]: # Step 1: Extract 'date' and 'FUTURE_RETURN' columns from the original data and add
data_processed_df['date'] = data['date'].values
data_processed_df['FUTURE_RETURN'] = data['FUTURE_RETURN'].values

# Step 2: Create prediction signal (use the best model to generate predictions)
backtest_df = data_processed_df.iloc[-len(X_test):].copy()
backtest_df['signal'] = blending_test_pred

# Step 3: Shift the signal to avoid data leakage

backtest_df['shifted_signal'] = backtest_df['signal'].shift(1)

# Step 4: Make sure 'backtest_df' is sorted by date in ascending order

backtest_df = backtest_df.sort_values('date')

# Step 5: Create the final 'backtest_df' with all columns needed for backtesting
backtest_df = backtest_df[['date', 'close', 'signal', 'shifted_signal', 'y', 'FUTUR

# Step 6: Check if 'backtest_df' is created successfully

backtest_df.head(20)

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 127/132

8/22/24, 11:05 PM Project-ML-CODE

Out[ ]: date close signal shifted_signal y FUTURE_RETURN

1360 2018-08-07 0.902385 1 NaN 0 -0.014659

1361 2018-08-08 0.870074 1 1.0 1 0.023158

1362 2018-08-09 0.920371 1 1.0 0 -0.006793

1363 2018-08-10 0.905275 1 1.0 0 -0.011584

1364 2018-08-13 0.879710 1 1.0 0 0.001340

1365 2018-08-14 0.882632 1 1.0 0 -0.025480

1366 2018-08-15 0.826973 1 1.0 0 -0.021666

1367 2018-08-16 0.780852 1 1.0 0 -0.020943

1368 2018-08-17 0.737236 1 1.0 1 0.015641

1369 2018-08-20 0.769129 1 1.0 1 0.041409

1370 2018-08-21 0.854883 1 1.0 0 -0.005644

1371 2018-08-22 0.842710 1 1.0 0 0.003759

1372 2018-08-23 0.850772 1 1.0 0 -0.014786

1373 2018-08-24 0.818943 1 1.0 1 0.040891

1374 2018-08-27 0.905661 1 1.0 0 -0.009166

1375 2018-08-28 0.885427 0 1.0 0 0.007122

1376 2018-08-29 0.901004 0 0.0 0 -0.007946

1377 2018-08-30 0.883500 0 0.0 0 -0.031173

1378 2018-08-31 0.815378 0 0.0 0 0.010649

1379 2018-09-03 0.837925 0 0.0 1 0.013254

In [ ]: # Define trading strategy logic

# Buy signal: Buy the stock when shifted_signal is 1.
# Sell signal: Sell the stock when shifted_signal is 0.

# Trading strategy decision tree

decision_tree_latex = r"""
\begin{array}{c}
\text{Start} \\
\downarrow \\
\text{Is } \text{shifted\_signal} = 1 \text{?} \\
\begin{array}{ll}
\text{Yes} & \quad \text{No} \\
\downarrow & \quad \downarrow \\
\text{Buy Stock} & \quad \text{Sell Stock} \\
\end{array}
\end{array}

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 128/132

8/22/24, 11:05 PM Project-ML-CODE

"""

display(Math(decision_tree_latex))

Start
↓
Is shifted\_signal = 1?
Yes No
↓ ↓
Buy Stock Sell Stock

In [ ]: # Calculate daily strategy return

# When shifted_signal is 1, calculate the return for the holding period; otherwise,
backtest_df['strategy_return'] = backtest_df['shifted_signal'] * backtest_df['FUTUR

# Check the strategy return in backtest_df

backtest_df[['date', 'close', 'shifted_signal', 'FUTURE_RETURN', 'strategy_return']

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 129/132

8/22/24, 11:05 PM Project-ML-CODE

Out[ ]: date close shifted_signal FUTURE_RETURN strategy_return

1360 2018-08-07 0.902385 NaN -0.014659 NaN

1361 2018-08-08 0.870074 1.0 0.023158 0.023158

1362 2018-08-09 0.920371 1.0 -0.006793 -0.006793

1363 2018-08-10 0.905275 1.0 -0.011584 -0.011584

1364 2018-08-13 0.879710 1.0 0.001340 0.001340

1365 2018-08-14 0.882632 1.0 -0.025480 -0.025480

1366 2018-08-15 0.826973 1.0 -0.021666 -0.021666

1367 2018-08-16 0.780852 1.0 -0.020943 -0.020943

1368 2018-08-17 0.737236 1.0 0.015641 0.015641

1369 2018-08-20 0.769129 1.0 0.041409 0.041409

1370 2018-08-21 0.854883 1.0 -0.005644 -0.005644

1371 2018-08-22 0.842710 1.0 0.003759 0.003759

1372 2018-08-23 0.850772 1.0 -0.014786 -0.014786

1373 2018-08-24 0.818943 1.0 0.040891 0.040891

1374 2018-08-27 0.905661 1.0 -0.009166 -0.009166

1375 2018-08-28 0.885427 1.0 0.007122 0.007122

1376 2018-08-29 0.901004 0.0 -0.007946 -0.000000

1377 2018-08-30 0.883500 0.0 -0.031173 -0.000000

1378 2018-08-31 0.815378 0.0 0.010649 0.000000

1379 2018-09-03 0.837925 0.0 0.013254 0.000000

Step 11.3: Calculate cumulative returns

Calculate the cumulative return for the strategy and the buy-and-hold strategy for
comparison.

In [ ]: # Calculate the cumulative return of the strategy

backtest_df['cumulative_strategy_return'] = (1 + backtest_df['strategy_return']).cu

# Calculate the cumulative return of the buy-and-hold strategy

backtest_df['cumulative_market_return'] = (1 + backtest_df['FUTURE_RETURN']).cumpro

# Visualize the cumulative returns

backtest_df[['date', 'cumulative_strategy_return', 'cumulative_market_return']].set
plt.show()

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 130/132

8/22/24, 11:05 PM Project-ML-CODE

Step 11.4: Calculate performance metrics

Compute key performance metrics such as cumulative return, max drawdown, annualized
return, and Sharpe ratio.

In [ ]: # Maximum Drawdown formula

display(Math(r'\text{Max Drawdown: } \text{MDD} = \max_{t} (\text{rolling\_max}(t)

# Annualized Return formula

display(Math(r'\text{Annualized Return: } R_{\text{annual}} = \left(1 + R_{\text{cu

# Annualized Volatility formula

display(Math(r'\text{Annualized Volatility: } \sigma_{\text{annual}} = \sigma_{\tex

# Sharpe Ratio formula

display(Math(r'\text{Sharpe Ratio: } S = \frac{R_{\text{annual}} - R_f}{\sigma_{\te

Max Drawdown: MDD = max(rolling\_max(t) − cumulative\_strategy\_return(t))

t

252
Annualized Return: Rannual = (1 + Rcumulative ) n −1

Annualized Volatility: σannual = σdaily × √252

Rannual − Rf
Sharpe Ratio: S = assuming Rf = 0
σannual

In [ ]: # Calculate cumulative return

cumulative_return = backtest_df['cumulative_strategy_return'].iloc[-1]

# Calculate maximum drawdown

rolling_max = backtest_df['cumulative_strategy_return'].cummax()
drawdown = rolling_max - backtest_df['cumulative_strategy_return']
max_drawdown = drawdown.max()

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 131/132

8/22/24, 11:05 PM Project-ML-CODE

# Calculate annualized return

annualized_return = ((1 + cumulative_return) ** (252 / len(backtest_df))) - 1

# Calculate annualized volatility

annualized_volatility = backtest_df['strategy_return'].std() * np.sqrt(252)

# Calculate Sharpe ratio (assuming risk-free rate is 0)

sharpe_ratio = annualized_return / annualized_volatility

performance_metrics = {
"Metric": ["Cumulative Return", "Max Drawdown", "Annualized Return", "Annualize
"Value": [cumulative_return, max_drawdown, annualized_return, annualized_volati
}

performance_df = pd.DataFrame(performance_metrics)

# Display the results

performance_df

Out[ ]: Metric Value

0 Cumulative Return 0.098988

1 Max Drawdown 0.255308

2 Annualized Return 0.072245

3 Annualized Volatility 0.250624

4 Sharpe Ratio 0.288259

file:///C:/Users/bob19/Desktop/Final Exam/Project-ML-CODE.html 132/132