MAINT Introduce use of set_output to output dataframes #683
Conversation
|
This is still draft so I did not merge. But feel free to undraft and merge. |
|
I think we should use |
The global setting raises an We can still set the output to be dataframe when creating the instances in the rest of the notebook, and use new instances with default input for the pipeline. |
| # %% | ||
| data_train_scaled = pd.DataFrame(data_train_scaled, columns=data_train.columns) | ||
| scaler = StandardScaler().set_output(transform="pandas") | ||
| data_train_scaled = scaler.fit_transform(data_train) |
There was a problem hiding this comment.
After the analysis, I would also some link to the documentation: https://scikit-learn.org/stable/auto_examples/miscellaneous/plot_set_output.html.
I would probably mention that we can set the output of a Pipeline using the sklearn.set_config function without going into details but instead providing delegating to the scikit-learn example.
There was a problem hiding this comment.
This comment may be relevant for Issue #675.
There was a problem hiding this comment.
I added a suggestion above to add the link right away without waiting for a PR dedicated to address #675.
|
In the notebook |
Co-authored-by: Guillaume Lemaitre <[email protected]>
|
+1 for |
|
Otherwise LGTM. |
| # %% | ||
| data_train_scaled = pd.DataFrame(data_train_scaled, columns=data_train.columns) | ||
| scaler = StandardScaler().set_output(transform="pandas") | ||
| data_train_scaled = scaler.fit_transform(data_train) |
There was a problem hiding this comment.
I added a suggestion above to add the link right away without waiting for a PR dedicated to address #675.
Co-authored-by: Olivier Grisel <[email protected]>
Pandas output with
set_outputAPI is available since v 1.2.This PR introduces such a nice feature to the MOOC.