evalml
evalml copied to clipboard
alteryx
EvalML is an AutoML library written in python.
Was reading the guide and noticed that the method names haven't been updated to `fit_and_transform_all_but_final` and `transform_all_but_final` [here](https://evalml.alteryx.com/en/stable/user_guide/component_graphs.html#Component-graph-methods). The release notes helped link me to the new names, which was...
Once https://github.com/alteryx/evalml/issues/1389 is finished, the `OrdinalEncoder` will exist but not actually be used within EvalML. We should integrate it. This will require a change to the OneHotEncoder, since when the...
Repro: ```python from evalml.pipelines.components import CatBoostClassifier import inspect assert inspect.getfullargspec(CatBoostClassifier.predict).args == ['self', 'X', 'y'] ``` This is confusing because it doesn't mean `cat_boost.predict(X, y)` will work. Also, the definition of...
Now that featuretools released a `NumericLag` primitive (https://github.com/alteryx/featuretools/pull/1797), all features created by `TimeSeriesFeaturizer` can be computed with dfs.
#1515 adds the initial version of the multicollinearity data check. However, it calculates the entire matrix, looking at two rows at a time for a O(n^2) algo. This issue tracks...
Right now, we access unique target values for classification problems in several ways: 1. `list(ww.init_series(np.unique(y)))` (`classification_pipeline.py`) 2. `unique_labels` (`confusion_matrix`) 3. LabelBinarizer / np.unique in `roc_curve` (slightly different than label encoding)...
Our `LogisticRegressionClassifier` only supports an `l2` [penalty](https://github.com/alteryx/evalml/blob/main/evalml/pipelines/components/estimators/classifiers/logistic_regression.py#L17), but the underlying sklearn implementation additionally [supports](https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LogisticRegression.html) `l1`, and `elasticnet`. I think we should add these regularizer options to our component hyperparameters. One...
This issue tracks writing our own implementation of dask's `as_completed` function to pull off jobs from the queue when they are done. This will be used to consolidate the "while-loop"...
From December Blitz: https://alteryx.quip.com/hifwAO0YHqul/Blitz-December-2020-Notes Notes are listed there: https://github.com/alteryx/evalml/pull/1589 Motivation: * While implementing the stacking ensemble method, many things were fitted (haha) for the stacking ensemble. For example, choosing to...