Warn about using scale, min_max_scale, etc.

[NicolasHug]

We have a bunch of preprocessing helpers like scale, quantile_transform, robust_scale etc.

While these can be useful, users misuse them and leak their training data, as recently illustrated in the mailing list.

I think we should:

1. Add a .. warning note in their respective docstrings indicating the dangers of using these and recommending pipelines + estimators instead
2. Stop referencing them from the UG, and only add them e.g. in the "See Also" sections, or even just in the API ref.

Right now the first entry of the Preprocessing guide is scale. I think it should be StandardScaler within a pipeline.

[NicolasHug]

@amueller thoughts on making step 1 a sprint issue? (I would provide a template)

[amueller]

lol when I saw the title I thought you wanted to warn, but saying something in the docs sounds good.

[thomasjpfan]

Yup lets make 1 a sprint issue. (Maybe even 2 if are daring)

[lucyleeow]

Does closure of #17402 mean this can be closed too?

[alfaro96]

If I am not wrong, we should also stop referencing them from the user guide. Maybe it would be interesting to open a new issue to handle this.

WDYT?

[lucyleeow]

Good point.

Maybe it would be interesting to open a new issue to handle this.

Sure or amend title of this one. Either works

[NicolasHug]

Let's indeed close this one, it'll be easier to keep track if we open another issue. Does one of you want to do it? Otherwise I will