ekamioka

Utils to format extraction of index and nunique of categorical columns

4

Implement auxiliary function to extract dict of indexes and unique values/dims ``` Usage: cat_dims = get_idxs_dims(df_train, cat_cols=categorical_features) clf = TabNetClassifier(cat_idxs=[x for x, _ in cat_dims.items()], cat_dims=[x for _, x in...

fix to the issue #90

6

Hey @Ironholds AlexCP said maybe I can help, so here is my PR in hope to solve it. I added new tests to ensure I am doing the thing (and...

This PR removes columns with constant values that will generate NaNs values durante the correlation matrix calculation. Therefore, I believe it is still required a discussion towards: - assess the...

AssertionError - assert version is not None

1

executed code: ``` >>> axgb = AutoXGB( ... train_filename="X_train.csv", ... output="output", ... test_filename="X_valid.csv", ... task="classification", ... idx=None, ... targets=["label"], ... features=['feat1', 'feat2', 'feat3', 'feat4', 'feat5'], ... categorical_features=None, ... use_gpu=False, ......

Support for GPT4all local models

1

First of all, superb project. I would like to contribute with this PR as the initial step required for performance experiments using open source models. Bests, K