Scribd
Upload
367 views40 pages

Modelo Multinivel Con R

This document provides an overview of using the statistical software R for multilevel modeling. It discusses the basic concepts of R including working with objects, basic operations, reading and saving data, and graphics. It also reviews using R for statistical analysis and introduces multilevel modeling in R. Specific functions and packages covered include lmer() for multilevel modeling, and libraries for reading data, generating graphs, and statistical analysis. The document is intended as an introduction for beginners to work with R for multilevel modeling.

Uploaded by

Jairo Cardona Giraldo
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
367 views40 pages

Modelo Multinivel Con R

This document provides an overview of using the statistical software R for multilevel modeling. It discusses the basic concepts of R including working with objects, basic operations, reading and saving data, and graphics. It also reviews using R for statistical analysis and introduces multilevel modeling in R. Specific functions and packages covered include lmer() for multilevel modeling, and libraries for reading data, generating graphs, and statistical analysis. The document is intended as an introduction for beginners to work with R for multilevel modeling.

Uploaded by

Jairo Cardona Giraldo
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 40

Multilevel Modeling with R

Spirin Nikita
Dorodnicyn Computing Center of the Russian Academy of Sciences 03.24.2010, Moscow

Packages covered

SAS MySQL Python Mathematica R

Agenda

R programming language and R Paradigm Basic operations in R Graphics with R Statistics with R Multilevel Models and ML with R

8 min. times 5 equals 40 min.

Overview

Free and commercialized GNU GPL R core team http://cran.r-project.org Interpreter

Concepts

Actions with in-memory objects function() function library

Basic Notation

Basic Notation

A-Z and a-z _ . 0-9 Case Sensitive

Basic operations in R

assign operator

Basic operations in R

ls() function

Basic operations in R

HELP

Reading Data

getwd() setwd() readtable() scan() read.fwf() ASCII Excel, SAS, SPSS, SQL-type databases

Reading Data

Saving Data
save.image()

Generating Data

Generating Data

Generating Data

Cartesian product

Generating Data
rfunc(n, p1, p2, ...)

Of course Matrices

Of course Matrices

Syntactic sugar

Graphics with R

Device paradigm

Window() Pdf() X11()

Graphics with R

Graphics with R

Legend for a graph

Graphics with R

Graphics with R

Graphics with R

Graphics with R

Graphics with R

Statistics with R

> library(stats)

Key operator ~ @model description operator y ~ model

Statistics with R

Statistics with R

Quiz y~x1+x2

y~I(x1+x2)

y ~ poly(x, 2)

Statistics with R

Multilevel Modeling with R

Multilevel Modeling with R

Why multilevel modeling?

Using all the data to perform inferences for groups with small sample size Predict an output for a new group Hierarchical models avoid overfitting effect of least squares regression Yields accurate measure of predictive uncertainty

Multilevel Modeling with R


fss = c(0,8,15,33,42,45,49,54,98,143,165,175,179,200) # include the library library(caTools) # read training and scoring data train <- read.csv("C:/Users/Spirinus/Desktop/Final Package/R/S_AUC_Train_1_7500.csv") score <- read.csv("C:/Users/Spirinus/Desktop/Final Package/R/S_AUC_Train_Test_7501_15000.csv") # data preparation train[train$Target == - 1, "Target"] <- 0 train$RowID = NULL

Multilevel Modeling with R


# build the model AUClogistic <- glm(Target ~ ., data=train[1:1000,fss+1], family=binomial(link="logit")) # get predictions on a scoring dataset test_scores <- predict(AUClogistic, type="response", score[1:1000,]) testY = score[1:1000,]$Target # calculate AUC colAUC(test_scores,testY)

Multilevel Modeling with R


lmer() library(matrix) Examples:

lmer(y ~ 1 + (1 | county))
lmer(y ~ x + (1 | county)) lmer(y ~ x + (1 + x | county))

Summary

How R works Basic objects in R R graphical capabilities R for statistical analysis Multilevel modeling in R

More Information

R for Beginners, Emmanuel Paradis, Institut des Sciences de l' Evolution Universite Montpellier II, F-34095 Montpellier cedex 05, France http://cran.r-project.org Data Analysis Using Regression and Multilevel Hierarchical Models, A. Gelman J.Hill

Acknowledgements

Thank you!

You might also like