Introductions

Neil Lund

Are you in the right room?

This is GVPT 628: Coding for Political Analysis

Instructor Info:

  • Neil Lund (nlund@umd.edu)

  • Office 1140C Tydings Hall (in the suite behind the glass doors)

  • Office Hours: Tu: 3:30 - 5:00, Thu: 11:00 - 12:30, MWF: by appointment

Materials

  • Access to computer with Python and R-Studio

  • Other materials provided through ELMS or Github

Course Goals

By the end of the semester you should:

  • Have a basic understanding of R (as well as some basic Python)

  • Understand how to get from messy real world data to “clean” analyses

  • Have some materials you can start using to build a portfolio

Most importantly: you should have a solid baseline to develop new skills on your own and collaborate with others.

Rough outline of the course

Pt 1: Basics

  • Loops and functions

  • Munging, aggregation, visualization

  • Survey analysis

  • Collaboration and presentation

Pt 2: Data collection, cleaning and project management

  • Automation and replication

  • Using APIs

  • Building/using a relational database, basic SQL

Pt 3: Analysis, Machine learning, Prediction

  • Dimensionality reduction

  • Inference, prediction, and model comparison

  • Working with text data

Grades

Homework and group assignment (60%)

  • More or less one a week. Submit on ELMS before class

  • Must include answers AND code (that works)

  • Will drop the lowest (if needed)

Participation and engagement (10%)

  • Not an attendance grade!

Final Project (30%)

  • Short analysis of a political science data set. Presented in last week of class. Memo due on final exam date.

General Guidelines for assignments

  • Your code should run with little-to-no modification from the person attempting to run it. Replicability is key.

  • I’ll generally leave some notes to highlight issues. “See me” typically sounds more ominous than it is.

  • Sometimes, we’ll do things the “hard way” just to illustrate how stuff works.

General Guidelines for assignments

AI Policy

In general, don’t.

AI generated search results are one minor exception. These are unavoidable, but they’re also fairly limited in what they tell you.

pictured: me, right now

pictured: me, right now

Set-up

  1. Creating an R project

  2. Turning off some settings

  3. Connecting to a Github repository

  4. Installing packages (if needed)