Comparing changes

Closes #69. ### Summary of Changes The `fit` method of classifiers/regressors now returns a new (fitted) classifier regressor. The receiver of the method call is not changed anymore. This is consistent with the methods on the `Table` class and other data containers. Furthermore, `fit` is now a pure function, which works better in notebooks and our own [execution strategy](https://arxiv.org/abs/2302.14556). --------- Co-authored-by: lars-reimann <lars-reimann@users.noreply.github.com>

…92) Closes #89. ### Summary of Changes We no longer raise an exception when `correlation_heatmap` is called on a table with some non-numerical columns. Instead, we simply don't show these columns in the created plot. --------- Co-authored-by: lars-reimann <lars-reimann@users.noreply.github.com>

Closes #93. ### Summary of Changes Rename `remove_outliers` to `drop_rows_with_outliers` in `Table`.

…#96) Closes #13. ### Summary of Changes Add a method `drop_columns_with_non_numerical_values` to the `Table` class to return a new table without columns with non-numerical values. --------- Co-authored-by: lars-reimann <lars-reimann@users.noreply.github.com>

Closes #10. ### Summary of Changes Add methods to `Table` to handle missing values: * `drop_columns_with_missing_values` returns a `Table` without the columns that have missing values. * `drop_rows_with_missing_values` returns a `Table` without the rows that have missing values. --------- Co-authored-by: lars-reimann <lars-reimann@users.noreply.github.com>

### Summary of Changes In `Table`, rename `keep_columns` to `keep_only_columns` to better match the semantics of the method.

Closes #94. ### Summary of Changes In the `Table` class: * Remove`list_columns_with_missing_values` * Remove `list_columns_with_non_numerical_values` * Mark `list_columns_with_numerical_values` as internal

Closes #7. ### Summary of Changes Previously, calling `drop_rows_with_outliers` on a `Table` that had at least one missing value in a numerical column cause the resulting table to be completely empty. This PR introduces two changes: 1. Missing values are never considered outliers. 2. Missing values are ignored when computing the standard deviation. --------- Co-authored-by: lars-reimann <lars-reimann@users.noreply.github.com>

## [0.6.0](v0.5.0...v0.6.0) (2023-03-27) ### Features * allow calling `correlation_heatmap` with non-numerical columns ([#92](#92)) ([b960214](b960214)), closes [#89](#89) * function to drop columns with non-numerical values from `Table` ([#96](#96)) ([8f14d65](8f14d65)), closes [#13](#13) * function to drop columns/rows with missing values ([#97](#97)) ([05d771c](05d771c)), closes [#10](#10) * remove `list_columns_with_XY` methods from `Table` ([#100](#100)) ([a0c56ad](a0c56ad)), closes [#94](#94) * rename `keep_columns` to `keep_only_columns` ([#99](#99)) ([de42169](de42169)) * rename `remove_outliers` to `drop_rows_with_outliers` ([#95](#95)) ([7bad2e3](7bad2e3)), closes [#93](#93) * return new model when calling `fit` ([#91](#91)) ([165c97c](165c97c)), closes [#69](#69) ### Bug Fixes * handling of missing values when dropping rows with outliers ([#101](#101)) ([0a5e853](0a5e853)), closes [#7](#7)

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Comparing changes

Open a pull request

Commits on Mar 27, 2023

This comparison is taking too long to generate.

Uh oh!