Fix Issue #2667

[JTPetter]

fixes https://github.com/jasp-stats/jasp-test-release/issues/2667

[koenderks]

What does this fix?

Edit: sorry didn't see. Did you test this? I'm at a conference without laptop so can't test but will happily approve if you did.

[JTPetter]

Yes, I tested it! Thank you

[koenderks]

Hmm interesting, can you explain this fix to me? Doesn't !identical(samples, "") also have length > 1? Sorry if this is a weird question but I don't have my laptop and am curious :p

[JTPetter]

Of course! "==" (or "!=") compares elements of an object element wise, so for instance entries in a vector one by one (but also matrices, dataframes etc.). So if your vector has length n it will return a boolean vector with length n. If you pass a boolean vector with length > 1 to a control statement in R, it will crash. I think this changed with R4.0, previously it would simply evaluate the first element of the boolean vector.

The identical function does a "deep" comparison, so it also checks whether the data type is the same. So the evaluated object will either be exactly an empty string or it returns FALSE. So the evaluation result is always a single boolean.