datamations is a framework for the automatic generation of explanation of the steps of an analysis pipeline. It automatically turns code into animations, showing the state of the data at each step of an analysis.

For more information, please visit the package website, which includes additional examples, defaults and conventions, and more.

You can install datamations from GitHub with:

# install.packages("devtools")
devtools::install_github("microsoft/datamations")

To get started, load datamations and dplyr:

A datamation shows a plot of what the data looks like at each step of a tidyverse pipeline, animated by the transitions that lead to each state. The following shows an example taking the built-in small_salary data set, grouping by Degree, and calculating the mean Salary.

First, define the code for the pipeline, then generate the datamation with datamation_sanddance():

library(datamations)
library(dplyr)

"small_salary %>% 
  group_by(Degree) %>%
  summarize(mean = mean(Salary))" %>%
  datamation_sanddance()

datamations supports the following dplyr functions:

• group_by() (up to three grouping variables)
• summarize()/summarise() (limited to summarizing one variable)
• filter()
• count()/tally