Ben J. Smith comments

Results 7 comments of


                                            Ben J. Smith

Feature request: geom_sf_label_repel(), geom_sf_text_repel()

Working on the same problem as @kendonB .

problem with dependency pyyawt on macos 12.4 on apple M1 arm processor

We can confirm this by trying to install `pyyawt` directly. ``` (test_rsHRF) localuser@localmachine % pip install pyyawt Collecting pyyawt Using cached pyyawt-0.1.1.zip (450 kB) Building wheels for collected packages: pyyawt...

Nested cross-validation example in docs seems to be not properly nesting CV as per scikit-learn documentation

Yes I considered forcing it not to do cross-validation. The only option that seemed viable was to use `model_selection.PredefinedSplit(test_fold)`. This would force it to use a consistent cross-validation. One could...

Nested cross-validation example in docs seems to be not properly nesting CV as per scikit-learn documentation

I hear the points that the Decoder is not designed to do principled tuning of different estimators. Having said that, it seems to be designed to test different parameters of...

Nested cross-validation example in docs seems to be not properly nesting CV as per scikit-learn documentation

> Another solution that may work to compare parameter impacts while avoiding recoding it would be to fit one Decoder for each parameter set.. I considered writing a CV algorithm...

Nested cross-validation example in docs seems to be not properly nesting CV as per scikit-learn documentation

@tbng do you think the suggested replacement in the OP makes sense for the documentation? I thought so when I posted it, but since hearing from others on here that...

nilearn.decoding.DecoderRegressor.fit attribute cv_params_ doesn't contain expected values

Hi, This is not the issue I thought it was, though it would still be nice to know why there's something called "beta" at the top level. Here's an example,...