Closed bearpelican closed 5 years ago
Check out this pull request on
You'll be able to see Jupyter notebook diff and discuss changes. Powered by ReviewNB.
I'm not sure what the use case is: if you have an empty tuple of things, the other transforms (that are very often applied over the tuple) will keep returning it empty. Could you provide an example of use?
Yup you are right. This was to fix an earlier bug that happened in MultiCategorize +OneHotEncode.
https://github.com/fastai/fastai_dev/pull/246 should be a better fix
_is_showable returns True on an empty list and returns early from Pipeline decode transformation. This isn't always ideal, because empty lists cannot be shown. We should give the other Transforms a chance to handle an empty list.
Ran into this error while decoding an empty MultiCategory class
Culprit:
all([])
returns true.