Apoc

New features
* The `ObjectSelector`, based on the `ObjectClassifier` for selecting objects of a given class. It produces an instance segmentation label image.
* The `ObjectMerger` has new options for features:
* `centroid_distance`,
* `mean_intensity_difference`,
* `standard_deviation_intensity_difference`,
* `area_difference`,
* `mean_max_distance_to_centroid_ratio_difference`

Potentially backwards breaking changes
* Internal refactoring in `ObjectClassifier` and `TableRowClassifier` may have caused backwards-compatibility breaking changes. This is not intentional. In case anyone experiences trouble, please create a github-issue and we'll fix it.

New features
* `ObjectMerger`

**Full Changelog**: https://github.com/haesleinhuepf/apoc/compare/0.10.0...0.11.0

Bugfixes
* Fix File not found error when training on folders (thanks to jo-mueller for reporting)
* Touching-neighbor-count may be weighted differently when training because of upstream changes in pyclesperanto-prototype 0.19.0

**Full Changelog**: https://github.com/haesleinhuepf/apoc/compare/0.9.0...0.10.0

New features
* enabled neighbor statistics for objec classification

Backwards compatibililty breaking changes
* apoc now depends on pyclesperanto_prototype >= 0.18.3

What's Changed
* Print classifier information when calling `str(classifier)` or `> classifier` on the ipython console by jo-mueller in https://github.com/haesleinhuepf/apoc/pull/31
* Feature importance by haesleinhuepf in https://github.com/haesleinhuepf/apoc/pull/30

Bug fixes
* Unknown features cause an exception now in the ObjectClassifier
* When older .cl files were read, this caused errors while loading because some fields were not filled with default values. Old .cl files should now work again,
* Added missing dependency: pandas

New Contributors
* jo-mueller made their first contribution in https://github.com/haesleinhuepf/apoc/pull/31

**Full Changelog**: https://github.com/haesleinhuepf/apoc/compare/0.8.0...0.8.1

Changed
* The default value for `num_ensembles` has been increased to 100 and is now in line with scikit-learn defaults. This higher number of trees makes results more reproducible.