Untidy datasets package

[srvanderplas]

There are lots of "tidy" data sets in packages, but it would be nice to have a package containing many different types of untidy data, to provide a convenient way to practice data wrangling skills.

Some potential source data sets:

• https://makingnoiseandhearingthings.com/2018/04/19/datasets-for-data-cleaning-practice/
• World Development Indicators (from a Kaggle thread about messy data practice)
• Quora thread
• Tidy Tuesdays data?

Documentation would probably need to explain the ways that each data set is untidy; vignettes could be used to show how to tidy things up.

[sctyner]

https://github.com/rfordatascience/tidytuesday/tree/master/data

[iqis]

Call it .....dirtyverse?

[kwstat]

There was a team that worked on this issue and created a package: https://github.com/kwstat/untidydata2