code

For replication code, see the research tab.

My other code can be accessed on GitHub, including the packages I’ve written (described below) and an R Guide, Python Guide, and Stata Guide.

R

R Guide to use with collaborators and research assistants to make R code consistent, easier to read, transparent, and reproducible.

tabulator efficiently tabulates and produces Stata tabulate-like output.

To install tabulator directly through R:

install.packages("tabulator")

tabulator includes the following functions:

• tab() efficiently tabulates based on a categorical variable, sorts from most common to least common, and displays the proportion of observations with each value, as well as the cumulative proportion.
• tabcount() counts the unique number of categories of a categorical variable or formed by a combination of categorical variables.
• quantiles() produces quantiles of a variable. It is a wrapper for base::quantile() but is easier to use, especially within data.tables or tibbles.

Python

Python Guide to use with collaborators and research assistants to make Python code consistent, easier to read, transparent, and reproducible.

get_files automates downloading files from a website using web scraping when you provide it with a url and the file extensions to scrape.

word2pdf automates Microsoft Word document to pdf conversion.

crop_eps crops .eps files; useful when you can’t get the cropping of a graph just right in your statistical software.

Stata

Stata Guide to use with collaborators and research assistants to make Stata code consistent, easier to read, transparent, and reproducible.

I’ve written a number of Stata packages that can be installed directly through Stata:

ssc install <package_name>, replace

exampleobs prints (randomly selected) example observations and optionally stores the values in a local macro. This is useful to explore possible values of a variable in your data set without being biased by the ordering of the data.

fiscal_impoverishment includes commands to estimate fiscal impoverishment (FI) and fiscal gains to the poor (FGP), which are measures of how much the poor benefit from or are hurt by the tax and transfer system from Higgins and Lustig (2016). Additional commands graph FI and FGP curves.

head prints the head observations (first observations in data set) and mimics the head() function in R and head command in Linux.

randomselect randomly selects observations and marks them with a dummy variable. It differs from sample in that it does not drop the non-selected observations from the data set, and that either individual observations or other units, defined by a variable in the data set, can be randomly selected.

tail prints the tail observations (last observations in data set) and mimics the tail() function in R and tail command in Linux. ​