[scikit-learn] GridSearchCV not working properly

[Guillaume Lemaître]

Hi Mujeebur,

On the top of the head, I don't recall any drastic changes in
GradientBoostingClassifier and GridSearchCV in the last version.
We start to continuously benchmark performance regressions and we did not
notice anything but it is possible that we missed it (or we do not test
with the sort of data that you have).

Please open an issue in GitHub with a reproducible minimal example and give
information regarding your platform and environment such that we can try to
reproduce:
https://github.com/scikit-learn/scikit-learn/issues

> Please fix this issue as soon as possible.

I would personally appreciate it if you could drop this type of statement
in your next post.

On Fri, 16 Apr 2021 at 14:38, Mujeebur Rahman <mujjub1997 at gmail.com> wrote:

> Dear Team,
>
> I really like to work with sklearn tools. But I've been facing issues ever
> since the latest version was released. It is about the GridSearchCV
> function of model_selection.
>
> With the previous version, I ran GradientBoostingClassifier with a set of
> hyperparameters and fed it to GridSearchCV, setting verbose to 5. The train
> data had around 800 rows and 100 columns. The model completed running
> within 2-3 minutes.
>
> With the latest version, I ran the same code without any changes. The only
> change is with the number of rows, where around 200 rows have been added
> extra. This time wheb I ran the code, it took nearly 20 minutes to
> complete. And the verbose steps also didn't show up.
>
> Please fix this issue as soon as possible. I really look forward to
> hearing from you.
>
> Thanks and Regards,
> Mujeebur Rahman
> Data Analyst
> _______________________________________________
> scikit-learn mailing list
> scikit-learn at python.org
> https://mail.python.org/mailman/listinfo/scikit-learn
>


-- 
Guillaume Lemaitre
Scikit-learn @ Inria Foundation
https://glemaitre.github.io/
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mail.python.org/pipermail/scikit-learn/attachments/20210416/979093d7/attachment.html>