evalml
evalml copied to clipboard
alteryx
EvalML is an AutoML library written in python.
As an EvalML dev, I would like test_imputer.py to be refactored in order to use a centralized dataframe for all tests. This dataframe should ultimately live in either conftest.py or...
As a [user/developer], I wish I could use EvalML to model with the new pandas nullable integer (Int64) type. Completion of this story involves adding appropriate coverage to the `imputer_test_data`...
Related to #2656. This issue tracks moving algorithm specific tests out of test_automl.py, test_automl_search_classification.py and test_automl_search_regression.py by mocking out next_batch and algorithm specific methods. This will improve the process of...
In conjunction with work being done with Innovation Days
In [`objective_base.py`](https://github.com/alteryx/evalml/blob/main/evalml/objectives/objective_base.py#L142-L143) we currently have logic to raise an error if `y_true` contains `NaN` or `inf`. However, this logic doesn't allow users to debug properly as there couple be multiple...
testing min_dependencies of min dependencies
- There is a regressor we could add to EvalML: - https://scikit-learn.org/stable/modules/generated/sklearn.neural_network.MLPRegressor.html - This came up because we want to properly treat cyclic categoricals - http://blog.davidkaleko.com/feature-engineering-cyclical-features.html - https://towardsdatascience.com/how-to-handle-cyclical-data-in-machine-learning-3e0336f7f97c - Cyclic...
Currently if no value for `y` is passed to the `TimeSeriesImputer`, the resulting output returns a `pd.Series` object with one None value due to [this](https://github.com/alteryx/evalml/blob/main/evalml/pipelines/components/transformers/imputers/time_series_imputer.py#L167). The behaviour should match our...
#2670 introduced `pydocstyle` and `darglint` packages. The `darglint` package specifically increases the runtime of our lint job by a few minutes. While we were okay with this addition, I suspect...
[This](https://github.com/alteryx/evalml/pull/3506) PR made `NoVarianceDataCheck` return only warnings, which limited the scope of the documentation of the Data Check Actions. The documentation for Data Check Actions should be expanded to include...