MONK's problems categorical features are wrongly represented as continuous/Integer

[phoeinx]

Describe the bug
"MONK's problems" features are all intended to be categorical by the paper creating the synthetic dataset.
In the UCI ML repository, they are all recorded as Integer. This leads to possible disadvantages for models trained on them.

To Reproduce
Steps to reproduce the behavior:

1. Go to https://archive.ics.uci.edu/dataset/70/monk+s+problems, and have a look at the variables table showing Integer as type for all features.
2. Open the MONK's problems competition paper freely available here: https://www.researchgate.net/publication/2293492_The_MONK's_Problems_A_Performance_Comparison_of_Different_Learning_Algorithms
3. Look at page 2, section 1.1. "The problem" to see that all features are actually categorical.

Expected behavior
Correct representation as categorical features for features a1,a2,a3,a4,a5,a6.

thank you!

[andrew-wang0]

Hi @phoeinx, thanks for the feedback. I'll defer to our librarians @markellekelly and @rlongjohn to confirm and adjust the dataset if necessary.