Correlation analysis
How are my variables related?

Overview

Before you start to create risk models, it is a good idea to visualize how the variables are related to one another. Risk models often involve correlated random variables, and exploring correlation between variables is an important part of exploratory data analysis. Learn how to measure the degree of linear dependency between two random variables using Python and the SciPy library. The lecture slides include a number of warnings related to dependency, causality and correlation in risk analysis.

This submodule is a part of the risk analysis module.

Course material

	Modelling correlations with Python Lecture slides (PDF) View on Slideshare
	Python notebook on correlation View Python notebook online Download Python notebook Run notebook in MyBinder Run notebook in Google Colab

In these course materials, applications are presented using the NumPy and SciPy libraries for the Python programming language.

Other resources

We recommend the following sources of further information on this topic:

• Analysis of the “pay for performance” (correlation between a CEO’s pay and their job performance, as measured by the stock market) principle at Freakonometrics

• Python notebook on a more sophisticated Bayesian approach to estimating correlation using PyMC

• Exploratory computing with Python, a set of Python notebooks on analyzing data using NumPy/SciPy