DOC: Correct misuse of term high-cardinality in docs.

[ebrassell]

• closes #xxxx
• tests added / passed
• passes black pandas
• passes git diff upstream/master -u -- "*.py" | flake8 --diff
• whatsnew entry

[jbrockmendel]

cardinality seems like a weird term to use here. can you suggest a clearer one?

[jorisvandenbossche]

It's indeed a very unclear term if you don't know it, but it's also commonly used in the machine learning community for this purpose (so it's certainly a correct term, at least with the fix here).
Since it is explain afterwards, I am fine with keeping it. But maybe switching it around (first the more easy explanation, then the technical term might be better)? Eg This is especially true for text data with only few unique values (low-cardinality data).

[ebrassell]

I agree that cardinality is the correct term here and certainly expect the target audience to be familiar with its proper usage, but I've reworded it as suggested for the benefit of those that might not be familiar with the term.

[WillAyd]

lgtm

[jbrockmendel]

LGTM

[jreback]

@gfyoung can you commit your suggestion and merge.

[gfyoung]

@jreback : This PR didn't allow for maintainers to make changes.

I will merge this and follow-up with the grammar fix.

[gfyoung]

Thanks @ebrassell !