-
Notifications
You must be signed in to change notification settings - Fork 332
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Support FeatureUnion for InvertableHashingVectorizer #16
Comments
Hm, do you mean it should be possible to pass FeatureUnion to InvertableHashingVectorizer, and it'll find all HashingVectorizers and apply itself to them? If so, it'd be also required to handle FeatureUnion recursively, and maybe add Pipeline support, because there can be other FeatureUnions in a FeatureUnion. I'd not put it directly to InvertableHashingVectorizer class; what about a helper function which returns a new FeatureUnion (or maybe a Pipeline)? It may be convenient for simple pipelines as well. |
Yeah, it looks hard to do in general, and we also need some sensible names for the different vectorizers (parts of union) to show them in the feature report. Right now I just construct I don't quite understand what the helper would do, sorry :) |
Aha, I finally understand what that helper could do: we already have some rudimentary support for FeatureUnion (meaning it works out of the box in simple cases), so that helper would just build a new feature union with hashing vectorizer replaces with inverting hashing vectorizer. |
FeatureUnion for text was implemented in #96, but it's still lacking unhashing support. Some work on it is in |
Just adding features from
.transformer_list
, possibly with prefixes, should be enoughThe text was updated successfully, but these errors were encountered: