Dealing with ill-formated characters about classifyr HOT 2 CLOSED

I will unlist all of the error messages and check for all.equal. If they are all equal, I will simply print the error message to console. Often, people don't realise that all of them are in the return variable.

Note that this problem actually is in randomForest package. The developer explains in October 2018:

My best guess is that at some point R converted the variable names to conform to its rules for symbol names (no spaces or dashes, or things like that—they get converted to dots). My suggestion is that you try not to use column names with those symbols to avoid such discrepancy. I will try and see if I can make randomForest() more accommodating with column names.

Could you send a request to [email protected] and ask for the problem to be fixed? I think I also got Mengbo at one time to e-mail and ask. Hopefully, if enough people badger Andy, he will use check.names = FALSE when data.frame are created as ClassifyR does.

from classifyr.

One representative error message printed to console if all cross-validations have same error message in version 2.9.1.

from classifyr.