Awesome

EDA (Exploratory Data Analysis)

Stata program that automates the generation of exploratory data analysis reports. The program classifies variables as categorical/continuous variables and uses this information to define what types of graphs and tables to use.

Installation

To install eda use the following command from Stata:

net inst eda, from(http://wbuchanan.github.io/eda)

Use case

Since Exploratory Data Analysis can take a substantial amount of time in addition to the time needed to clean/prep data, this is intended to be used as a program that would be called at the end of the workday/overnight to produce permutations of univariate and bivariate visualizations and tables. Then instead of spending time coding myriad possible combinations of variables to examine, a researcher could browse through a PDF generated through LaTeX while the computer does the work of compiling the results for them.

Usage

You need a dataset open in memory to use eda. There are two required options, output for the naming of the output files and root to tell eda where to store the output.

sysuse auto // load the data
eda, o("eda-report") root("./") // use current working directory

Alternatively, you can restrict the variables to use for eda with a varlist:

clear
sysuse auto
eda price mpg weight, o("eda-report-small") root("./") 

Dependencies

This program requires a few other user-written programs to execute:

tuples spineplot estout brewscheme

You can install these dependencies using:

ssc install tuples
ssc install spineplot
ssc install estout
net install brewscheme, from("https://wbuchanan.github.io/brewscheme")

You can find information about these packages using:

ssc d tuples
ssc d spineplot
ssc d estout