Fix FR3 Exception

[F-said]

Describe the bug
Must raise an exception when data being filtered is invalid. Check in the negative rather than the positive ("if data is NOT raw data" vs "if data is None, if data is NaN, if data is...")

Expected behavior
Exception is raised

Screenshots

[F-said]

Consider what errors could occur while using the filter function:

• user tries to use Epochs into filter (is this really an error?) TypeError
  • Need feedback on. Why would this happen?
• user tries to use silly data (ex: pd dataframe) w/filter TypeError
  • hard stop maybe? don't try to preprocess garbage data. Need feedback on this as well.
• ValueError: Stop bands are not sufficiently separated ValueError
  • bad params, don't try to filter, but continue w/preprocess

If we assume that filter_feature only runs w/in pipeline, we can safely ignore these exceptions. However, say a savvy researcher wants to dig around this code and use the features in their own forked branch, then appropriate error handling would be nice.