docs: fit transform addition

[davidbp]

I am missing a method that might allow fitting a transformation and returning the transformed result in a single call.

Here there is an example from sklearn that shows the usage of fit_transform

>>> from sklearn.feature_extraction.text import CountVectorizer
>>> corpus = [
...     'This is the first document.',
...     'This document is the second document.',
...     'And this is the third one.',
...     'Is this the first document?',
... ]
>>> vectorizer = CountVectorizer()
>>> X = vectorizer.fit_transform(corpus)
>>> vectorizer.get_feature_names_out()
array(['and', 'document', 'first', 'is', 'one', 'second', 'the', 'third',
       'this'], ...)
>>> print(X.toarray())
[[0 1 1 1 0 0 1 0 1]
 [0 2 0 1 0 1 1 0 1]
 [1 0 0 1 1 0 1 1 1]
 [0 1 1 1 0 0 1 0 1]]

Note that you could just think that fit_transform simply has fit and transform inside as two function calls, but this would require iterating over the data twice (one for each function call).

One benefit of fit_transform is that it can iterate only once over the data and generate the transformed data while it is iterating over it.

If no specific efficient fit_transform is implemented it could be just sintactic sugar for calling fit!(transformer, X) and then transform(transformer, X)

[ablaom]

Sorry for my late response, and thanks for bringing this up.

Yes, I understand that there is a use-case for fitting and transforming in one go; as you say one can avoid extra computation/allocation.

Currently my head is around reducing methods as much as possible, and so I will come back to this but keep it in mind.

[ablaom]

closing in favour of #18