ENH Add parameter return_X_y to make_classification

[SuccessMoses]

Reference Issues/PRs

Fixes #16532

What does this implement/fix? Explain your changes.

• add parameter return_X_y to make_classification

Any other comments?

The dataset returned by load_iris is a Bunch, which is more descriptive. #16532 proposes same.

from sklearn.datasets import load_iris

data = load_iris()
print(data.DESCR)  # Prints a description of the Iris dataset

[github-actions]

✔️ Linting Passed

All linting checks passed. Your pull request is in excellent shape! ☀️

_{Generated for commit: a5754db. Link to the linter CI: here}

[SuccessMoses]

@adrinjalali could you review this?

[adrinjalali]

Please have a look at fetch_openml on how to document return_X_y and the documentation of the return values.

[OmarManzoor]

Thanks for the PR @SuccessMoses
I added a few comment

[SuccessMoses]

@OmarManzoor Thanks for the review. I am working on it

[OmarManzoor]

LGTM. Thanks @SuccessMoses

[OmarManzoor]

CC: @adrinjalali

[adrinjalali]

Otherwise LGTM.

[adrinjalali]

this isn't modifying the existing variable, it's allocating a new chunk of memory, and the nameweights refers to that new chunk. So the original data passed to this function is never changed. Therefore this change in this PR is unnecessary.

[SuccessMoses]

I did not intend to modify the variable weights passed by the user this is why I created new variable weights_

[adrinjalali]

You won't be changing the original variable, that's now how python works 😉

Investigate this example:

import numpy as np

a = np.array([1, 2, 3, 4, 5, 6, 7, 8, 9, 10])

def f(a):
    a = a + 1
    return a

print(f(a))
print(a)

[SuccessMoses]

this isn't modifying the existing variable, it's allocating a new chunk of memory,

Thank you @adrinjalali for the correction. make_classification returns a bunch object which also contains a dictionary of the original value of parameters like n_samples, n_features and weights that was used to generate the data.

https://github.com/SuccessMoses/scikit-learn/blob/test/sklearn/datasets/_samples_generator.py#L356-L383

reassigning the variable weights will change its content. Is there a way to work around this?

[adrinjalali]

I see, I'd missed that