Scikitlearn is the de facto home for all kinds of modeling algorithms. It has a plethora of
algorithms but still one thing that seems to be missing is the implementation of LogisticRegression where we can have p-values.

It would be great if we have something like model.p_values_ attribute for the Logistic Regression Models.

I know that there is another statistical library statsmodels which provides p_values, but a lot of programmers use sklearn and they build models based on this library. It is somewhat
inconvenient to use statsmodels just to get p-values and run other models such as Random Forest in sklearn.

Afterall, the API of statsmodels and sklearn are quite different. sklean is trend setter and
most people feel comfortable with sklearn API, however, statsmodels follows R-programming API and they are quite different.

In conclusion, It would be great if sklearn provides p-values for linear models.

I am eagerly waiting for the implementation in future versions of sklearn.