This project provides one Python 3 module, nrsci. It's used to tranform New Relic data (anything that can be queried with NRQL) into Pandas data frames.

Sometimes we want to perfrom complex calculations with our data and NRQL isn't able cover our use cases. For these cases we have nrsci, that allows us to download data from New Relic in any arbitrary size, not limited to 2000 results, and convert it into data frames so we can apply any data science technique.

First clone this repo, and then install the nrsci module with pip3:

$ pip3 install git+file:///path/to/cloned/repo/NRSci

NRSci makes use of the NerdGraph API, so you will need to set up a user key.

You can use nrsci to load data from your New Relic account and perform complex calculations using Python:

from nrsci import NerdGraph, Mapper
import pandas as pd
import datetime as dt

# Set up NerdGraph access
ng = NerdGraph("YOUR ACCOUNT ID", "YOUR USER KEY", NerdGraph.Endpoint.US)

# Configure mapper, to obtain events of type 'MyEventType' from New Relic, since 60 minutes ago until now
mapper = Mapper(ng, "MyEventType").since(dt.datetime.now() - dt.timedelta(minutes = 60)).until(dt.datetime.now())

# Request data and return a Pandas DataFrame
df = mapper.request()

When creating the Mapper you can specify multiple event types separated by commas:

mapper = Mapper(ng, "MyEventType_1", "MyEventType_2", "MyEventType_N")

You can also avoid since() and until(), their default values are 60 minutes ago and now, respectively.

You can also specify a SELECT statement, exactly like you would do in NRQL:

mapper = Mapper(ng, "MyEventType").select("actionName, round(timeSinceLoad / 60.0)")

If not specified, the default value is "*".

NRSci is licensed under the Apache 2.0 License.