Zebin YANG

Hi @parunach, I also encountered this error with an old version of pip. You may need to upgrade your pip to the latest version via "pip install pip -U", and...

Closed as starting from PiML-0.5.1, Mac M1 is supported with Python 3.8, 3.9, and 3.10.

Hi @HamedMZNN , I don't understand your request. Can you elaborate more about "different Feature Selectors and different Classifiers"? (better with some demo codes)

Hi @HamedMZNN , You may write some scripts to do such kinds of feature engineering and model training. Just register the model and data into PiML using the demo at...

Hi @rfox12-edu, can you share a reproducible script here?

Hi @yigitcancomlek， Sure, the 10MB limit is due to the widgets interface. PiML can handle larger data by pandas dataframe, see the user guide in https://selfexplainml.github.io/PiML-Toolbox/_build/html/guides/data/data_load.html#external-dataset.

Hi @Chenchaoyou, Thanks for mentioning that, and the contents in user guide 6.6 were not updated to the example figures. When the worst-sample ratio is 10%, only the worst-10% samples...

The PiML built-in models follow the sklearn style, and the sklearn's grid-search API can be used directly. For example, to tune the L1 and L2 regularization strengths, the following codes...

That makes sense. We may provide an enhanced leaderboard panel that adopts the model ranking and segmented metrics in future releases.

Thank you for this valuable suggestion. Currently, the residual plot is based on full samples. We will consider adding the option to visualize only a subset of samples to be...