✔️ Linting Passed

All linting checks passed. Your pull request is in excellent shape! ☀️

_{Generated for commit: 15acf02. Link to the linter CI: here}

thanks for the comments.

1. I will use RobustScaler instead of StandardScaler in both python examples.
2. There is another problem: the iterative imputer in the example plot_missing_values.py uses estimator=BayesianRidge which also requires features on a similar scale for its best performance (due to ridge's L2 optimization of weights), so I will change this as well (and update the issue description).

Here is the summary of the updates. In plot_missing_values.py:

• re-phrasing, typos
• the same order throughout the file: first, process the diabetes data, then process the california data
• using a single evaluation function for both full and imputed data (to remove code repetitions)
• removed the argument missing_values=np.nan in the calls to all imputers
• changed order: first, compute scores for ImputeByZero, then for ImputeByMean
• as suggested, using RobustScaler instead of StandardScaler for the California dataset (and no scaling for the diabetes dataset)
• in the California dataset, use RobustScaler before imputing with BayesianRidge (which is the default) to put the features on the same scale
• the labels on the diagram do not contain "scaled", because the scaling is applied only to the California dataset

Changes in file plot_iterative_imputer_variants_comparison.py:

• using RobustScaler instead of StandardScaler as was suggested
• using the single function compute_score_for for computing the scores for both full and missing-values datasets
• using scaling for all evaluations, because the final target estimator is BayesianRidge which is affected by different-scale features (due to regularization)
• modified call parameters of Ridge, k-NN regressor, and RandomForestRegressor, to improve computational stability during iterative imputation, and increased the number of iterations
• removed median simple imputation (does not add anything interesting in this context)

Looks good to me. What do you think @ArturoAmorQ?