Changeforest

---------------------

**Bug fixes:**

- The pseudo-permutation-test now correctly skips the first and last `minimal_relative_segment_length * n` observations when calculating the permuted maximal gains.

**Other changes:**

- The first three elements of the `result.optimizer_result.gain_results` returned by the two-step search are no longer sorted by their maximal gain.

---------------------

**Breaking changes:**

- The parameters `random_forest_mtry` and `random_forest_n_trees` of `Control` have been renamed to `random_forest_max_features` and `random_forest_n_estimators`.
- The default value for `random_forest_max_features` now is `floor(sqrt(d))`.

**New features:**

- The parameter `random_forest_max_features` now can be supplied with a fraction `0 < f < 1`, an integer `i>=1`, `None` (Python, Rust) / `NULL` (R) and `"sqrt"`. Then, for each split, repsectively `floor(f d)`, `i`, `d` or `floor(sqrt(d))` features are considered.

**Other changes:**

- Bump `biosphere` dependency to 0.3.0

---------------------

**Other changes:**

- Bump `biosphere` dependency to 0.2.2.

---------------------

**Other changes:**

- The default value for `Control.minimal_gain_to_split` is now `log(n_samples) * n_features / n_samples`,
motivated by the Bayesian information criterion (BIC). This is relevant for `change_in_mean` only.

---------------------

**Other changes:**

- The R-package now makes use of the latest version of `libR-sys`, enabling compilation for Apple silicon on `conda-forge` (86).

**Bug fixes:**

- Fixed a bug where passing `Control()` to `changeforest` in the Python package overwrote the default value for `random_forest_max_depth` to `None`. Default values for `Control` in the Python package are now `"default"` (87).

---------------------------

**Bug fixes:**

- Upgrade `biosphere` to `0.2.1` fixing a bug in `RandomForest`.

**Other changes:**

- New parameter `model_selection_n_permutations`.