changeforest 0.6.1

0.6.1 - (2022-04-06)

Bug fixes:

• Fixed a bug in the Python package when passing random_forest_max_features='sqrt' to Control.

changeforest 0.6.0

0.6.0 - (2022-03-17)

Breaking changes:

• The default value for model_selection_n_permutations is now 199.
• The default value for model_selection_alpha is now 0.02.
• The default value for minimal_gain_to_split, use in the change_in_mean setup, is now log(n) * (d + 1), motivated by the BIC and [1].
• The value for minimal_gain_to_split no longer gets automatically multiplied by the number of samples n.

[1] Yao, Y.-C. (1988). Estimating the number of change-points via Schwarz’ criterion. Statist. Probab. Lett. 6 181–189. MR0919373

changeforest 0.5.1

0.5.1 - (2022-03-16)

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.

changeforest 0.5.0

0.5.0 - (2022-03-15)

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

changeforest 0.4.4

0.4.4 - (2022-02-22)

Other changes:

• Bump biosphere dependency to 0.2.2.

changeforest 0.4.3

0.4.3 - (2022-01-29)

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.

changeforest 0.4.2

0.4.2 - (2022-01-21)

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).

changeforest 0.4.1

0.4.1 - (2022-01-13)

Bug fixes:

• Upgrade biosphere to 0.2.1 fixing a bug in RandomForest.

Other changes:

• New parameter model_selection_n_permutations.

changeforest 0.4.0

0.4.0 - (2022-01-11)

New features:

• changeforest now uses random forests from biosphere. This should be faster than smartcore used previously and supports parallelization.

changeforest 0.3.0

0.3.0 - (2021-12-15)

New features:

• Implemented trait Display for BinarySegmentationResult. In Python str(result) now prints a pretty display (#77).

Other changes:

• The TwoStepSearch algorithm now only uses valid guesses from split_candidates (#76).

Bug fixes:

• (R only) The R6 class Control now gets correctly exported (#79).