Tutorial Overview

• Introduce attendees not familiar with Principal Component Analysis (PCA) to the commonly used dimension reduction approach.
• Focus on real data analysis with R and Python.

Tutorial Prerequisites

Attendees are expected to have a working knowledge of statistics or linear algebra and R or Python.

Tutorial Outline

• Motivation for PCA
• Overview of PCA (e.g. checking of assumptions, deciding how many components to use, etc.)
• Two applications of PCA
• Extensions of PCA

Running the Examples

Software Prerequisites

Installation of R/Python and the following R/Python packages:

• R: lattice, blockcluster (optional), RTextTools (optional), RJSONIO (optional), httr (optional)
• Python: matplotlib, numpy, os, pandas, scipy, sklearn

If scraping Meetup data from scratch:

• Get an API key from Meetup.com: https://secure.meetup.com/meetup_api/key/
• Save it as a string in the file 'meetup.apikey' (located in the 'Code' folder)

Running the examples

• Clone the repository and navigate to 'Code' folder in the Terminal.
• Example 1: Run the R/Python code as follows:

# In R:
source("Example1.R")

# In Python:
execfile("Example1.py")

• Example 2: Run the R/Python code as follows:

# In R:
source("Example2.R")

# In Python:
execfile("Example2.py")