tldr: I created an open dataset for the explicit practice of data munging. Feel free to use it in assignments, but do mention where you got it from (CC-by-4.0). Also unicorns are awesome.
Find the dataset at: https://github.com/RMHogervorst/unicorns_on_unicycles
Data munging / cleaning / engineering
At work I was working with a two excel files that were slightly different but could be combined into 1 dataset. This is very typical for day to day cleaning operations that analysts and data scientists do (statisticians too).
I realized cleaning, joining and enriching is something that statistics classes just take for granted. But if a student only works with perfectly prepared data, they are unable to work with real world data. Because the real world is someone handing you an excel file with weird values and beautiful colors, that you cannot use in your work. Or it is a webscraping exercise where some of the pages are missing and people can’t seem to spell right. Some people say 80% of a data scientists work is cleaning data, so let us teach students how to do that effectively. I made this dataset in R, but it does not really matter what tool you use to read and clean this set.
What is the dataset about?
All good datasets have a story, this dataset is thought to have been recorded by an amateur scientist, a natural philosopher by the name of Rudolphus in the 17th century in The Netherlands. This scientist recorded the annual population of unicorns in western Europe over a century and also recorded the sales of unicycles in that same time period. Although not much of the accompanying text remains of the original documents, what we can read is the tables and the idea that Rudolphus thought there was some sort of relationship between uncorns and unicycle sales.
what kind of tasks would a student have to do?
There are 2 files, that contain a total of 3 tables. The tables can be connected to each other with full joins on countryname and year (To make it more difficult, you could remove one row in one of the sets).
a student should:
- read in the data from excel
- recognize that one file contains 2 tables
- realize that countrynames are slightly different in 1 file
- join the files together
- create 1 ‘long’ tidy datafile (country, year, value1, value2, value3)
During analysis:
- realize that some years are missing
- perhaps impute, reason for reasons of missingness (disease epidemic in Austria, others are missing at random)
- find some relation between unicycles and unicorns in the 17th century in western Europe
what can I do with this set?
I’ve released the set under the creative commons cc-by-4.0 license. That means you may remix, share, use commercially, and modify the dataset. I just want you to mention where you got the dataset from. If you think the units are stupid, change it. if you think the assignment is too difficult, make it easier. Really, you can use it any way you want.
I hope this set will help in teaching students the art of data munging. I also thought it was very funny to have unicorns on unicycles, but searching the internet shows I was hardly the first one to think of that pun. Ah well, just goes to show that it was a good one.
Find the dataset at: https://github.com/RMHogervorst/unicorns_on_unicycles
Happy coding!