Jérémie du Boisberranger

@glemaitre this is the plot that we talked about irl, with a histogram distribution of the residuals on the side that I found interesting  I don't think it needs...

I agree that it requires to be careful when interpreting these plots, but it's the case for everything. I don't think it's enough to just discard it. Regarding the residuals...

I don't know how to solve this debate :( maybe some @scikit-learn/core-devs can help us decide ? To summarize, this PR proposes a new display object for regression which can...

I'm also strongly in favor of using cnp.float32/64_t. It reduces confusion a lot

I encountered the same issue. The 5 bins test was added in 0.20.3. I agree that the test might be a bit ambiguous because it involves empty clusters and ties...

I'm not sure I can help you. Basically we can't guarantee that the private api won't change in the future. I'd advise you to only rely on the public api...

Note that https://github.com/scikit-learn/scikit-learn/pull/25810 supersedes this PR with the solution decided in https://github.com/scikit-learn/scikit-learn/issues/25572

#1359 added a warning to report pickling errors that occur in the caching mechanism, which will be turned into an error in version 1.5. It also added a warning (`CacheWarning`)...

Here is the output of conda list scikit-learn and daal ``` Name Version Build Channel scikit-learn 0.19.1 np114py36_35 intel daal 2019.0 intel_117 intel pydaal 2019.0.0.20180713 np114py36_0 intel ``` After upgrading,...

We won't have time to finish the review on this one before the 1.2 release. Moving it to 1.3