Data science

Visualizing varying effects' posteriors with joyplots

You should probably use ‘joy plots’ to visualize varying effects’ posteriors, because they look great.

Quantitative literature review with R: Exploring Psychonomic Society Journals, Part II

In this tutorial, I’ll show how to use R to quantitatively explore, analyze, and visualize a research literature, using Psychonomic Society publications. This post directly continues from part I of Quantitative literature review with R. Please read that first for context. Part I focused on data cleaning and simple figures, but here we will look at relational data by visualizing some network structures in the data.

Quantitative literature review with R: Exploring Psychonomic Society Journals, Part I

In this tutorial, I’ll show how to use R to quantitatively explore, analyze, and visualize a research literature, using Psychonomic Society’s publications

Scraping Statistical Modeling, Causal Inference, and Social Science

Warning: 'theme_blog' is deprecated. ## Use 'theme_vmisc' instead. ## See help("Deprecated") Hi everybody!

Today I’ll share some tips on elementary web scraping with R. Our goal is to download and process an entire Wordpress blog into an R data frame, which can then be visualized and analyzed for fun and discovery. We’ll scrape andrewgelman.com, the home of “Statistical Modeling, Causal Inference, and Social Science”. This is a very popular statistics and social science blog, whose main author, Andrew Gelman is a famous statistician and political scientist, and author of such classic holiday thrillers as Bayesian Data Analysis and Data Analysis Using Regression and Multilevel Models.

GitHub-style waffle plots in R

I have a little python script on my work computer that tracks hours I’ve spent in the office (assuming that those hours begin with a computer login, and end in a computer shutdown–sadly this is almost always true.) I don’t really do anything with the information, and it is really a remnant of my time as a contract worker. However, for shared or billed projects, such automation of data gathering can be very valuable.