Skip to content

Releases: henrikbostrom/crepes

crepes 0.7.1

21 Sep 15:31
acf8c68
Compare
Choose a tag to compare

v0.7.1 (21/09/2024)

Features

  • The calibrate methods of the classes WrapClassifier and WrapRegressor now take an additional argument seed, for setting the state of the random number generator. This allows for predict_p and predict_set of the former class and predict_int and predict_cps of the latter class as well as evaluate for both classes to become deterministic. The methods predict_p, predict_set, predict_int ,predict_cps and evaluate of these classes also have an argument seed, which can be used to over-ride any setting by the calibrate method. In addition, the corresponding methods of the classes ConformalClassifier and ConformalPredictiveSystem also include the argument seed for the same purpose. (The methods of ConformalRegressor currently contains no stochastic components and there is hence no need for a seed.) Thanks to @egonmedhatten and @tuvelofstrom for suggesting this extension.

  • The predict_p methods of the classes ConformalClassifier and WrapClassifier now takes an optional argument smoothing to allow for generating both smoothed and non-smoothed p-values (default: smoothing=True).

  • The default value for the parameter smoothing of the predict_set and evaluate methods has been changed to True.

  • The class DifficultyEstimator in crepes.extras now includes a parameter f for providing a function to compute the difficulty estimates.

Fixes

  • The documentation for the class MondrianCategorizer in crepes.extras has been corrected.

crepes 0.7.0

27 Jun 15:00
e11ecd3
Compare
Choose a tag to compare

v0.7.0 (27/06/2024)

Features

  • The class MondrianCategorizer was added to crepes.extras, for generating categories to be used by Mondrian conformal classifiers, regressors and predictive systems. See the documentation for the interface to objects of the class through the fit and apply methods.

  • The class WrapRegressor has been updated so that the method calibrate allows for specifying an (optional) difficulty estimator and (optional) Mondrian categorizer, which are used both for calibration and making predictions, instead of requiring that difficulty estimates and Mondrian categories are provided separately for these tasks. The methods predict_int, predict_cps and evaluate no longer require sigmas and bins to be provided for normalized and Mondrian conformal regressors and predictive systems. Thanks to @tuvelofstrom for suggestions along these lines.

  • The class WrapClassifier has been updated so that the method calibrate allows for specifying an (optional) Mondrian categorizer, which is used both for calibration and making predictions. The methods predict_p, predict_set and evaluate no longer require bins to be provided for Mondrian conformal classifiers.

Fix

  • Label vectors represented by pandas.Series are converted to NumPy arrays to avoid indexing issues. Thanks to @valeman for pointing this out.

crepes 0.6.2

02 Feb 15:19
8d483f3
Compare
Choose a tag to compare

v0.6.2 (02/02/2024)

Fixes

  • Fixed deprecated code for checking if an array is non-empty in the ConformalPredictiveSystem class. Thanks to @tuvelofstrom for pointing this out.

crepes 0.6.1

21 Aug 07:46
68b4e40
Compare
Choose a tag to compare

v0.6.1 (21/08/2023)

Features

  • The function margin for computing non-conformity scores for conformal classifiers has been added to crepes.extras.

Fixes

  • Fixed a bug in the DifficultyEstimator class (in crepes.extras), which caused an error when trying to display a non-fitted object. Thanks to @tuvelofstrom for pointing this out.

  • Fixed an error in the documentation for the function hinge.

  • The Jupyter notebooks crepes_nb_wrap.ipynb and crepes_nb.ipynb have been updated to illustrate the new margin function.

crepes 0.6.0

28 Jun 19:48
040af7a
Compare
Choose a tag to compare

v0.6.0 (28/06/2023)

Features

  • The classes ConformalClassifier and WrapClassifier have been added to crepes, allowing for generation of standard and Mondrian conformal classifiers, which produce p-values and prediction sets. The calibrate method of WrapClassifier allows for easily generating class-conditional conformal classifiers and using out-of-bag calibration. See the documentation for the interface to objects of the class through the calibrate, predict_p and predict_set methods, in addition to the fit, predict and predict_proba methods of the wrapped learner. The method evaluate allows for evaluating the predictive performance using a set of standard metrics.

  • The function hinge for computing non-conformity scores for conformal classifiers has been added to crepes.extras.

Fixes

  • The class Wrap has changed name to WrapRegressor and the arguments to the calibrate method of this class have been changed to be in line with the calibrate method of WrapClassifier.

  • The Jupyter notebooks crepes_nb_wrap.ipynb and crepes_nb.ipynb have been updated and extended

crepes 0.5.1

22 Jun 15:09
28977fb
Compare
Choose a tag to compare

v0.5.1 (22/06/2023)

Fix

  • Fixed a bug in the evaluate method of ConformalPredictiveSystem, which caused an error when using CRPS as a single metric, i.e., when providing metrics=["CRPS"] as input. Thanks to @Zeeshan-Khaliq for pointing this out.

crepes 0.5.0

02 Jun 13:19
a572897
Compare
Choose a tag to compare

Feature

  • The full cpds matrix is calculated only if requested to be output (return_cpds=True) by the predict method of ConformalPredictiveSystem or if the set of metrics include "CRPS" for the evaluate method. This allows large test and calibration sets to be handled without excessive use of memory in other cases. Thanks to @christopherjluke and @SebastianLeborg for highlighting and discussing the problem.

Fixes

  • Default values for mandatory arguments for the methods fit, predict and evaluate methods of ConformalRegressor and ConformalPredictiveSystem, as well as the function binning in crepes.extras, are no longer provided

  • y_min and y_max correctly inserted for all percentiles

  • The evaluate method for ConformalPredictiveSystem fixed to work correctly even if CRPS not included in metrics, and if all test objects belong to the same Mondrian category

  • Incorrect values for percentiles will render an error message

crepes 0.4.0

16 May 13:38
e8d2709
Compare
Choose a tag to compare

v0.4.0 (16/05/2023)

Feature

  • The class Wrap has been added to crepes, allowing for easily extending the underlying learner with methods for forming, and making predictions with, conformal regressors and predictive systems. See the documentation for the interface to objects of the class through the calibrate, predict_int and predict_cps methods, in addition to the fit and predict methods of the wrapped learner.

Fixes

  • A Jupyter notebook crepes_nb_wrap.ipynb has been added to the documentation to illustrate the use of the Wrap class.

  • The output result array of a conformal predictive system is converted to a vector if the array contains one column only.

  • The documentation has been updated and now includes links to classes and methods.

  • crepes.fillings has been renamed to crepes.extras

crepes 0.3.0

11 May 13:09
7628e1e
Compare
Choose a tag to compare

Features

  • The class DifficultyEstimator was added to crepes.fillings, incorporating functionality provided by the previous functions sigma_knn, sigma_knn_oob, sigma_variance, and sigma_variance_oob, which now are superfluous and have been removed from crepes.fillings. See the documentation for the interface to objects of the class through the fit and apply methods.

  • An option to normalize difficulty estimates, by providing scaler=True to the fit method of DifficultyEstimator, has been included.

Fixes

  • The Jupyter notebook crepes_nb.ipynb has been updated to incorporate the above features

  • The documentation of the crepes package and the crepes.fillings module has been updated with links to source code, additional examples and notes.

crepes 0.2.0

28 Apr 08:53
5e2cb52
Compare
Choose a tag to compare

Features

  • Modified sigma_knn to allow for calculating difficulty in three ways; using distances only, using standard deviation of the target and using the absolute residuals of the nearest neighbors.
  • Added sigma_knn_oob in crepes.fillings
  • Renamed the performance metric efficiency to eff_mean (mean efficiency) and added eff_med (median efficiency) to the evaluate method in ConformalRegressor and ConformalPredictiveSystem
  • Added warning messages for the case that the calibration set is too small for the specified confidence level or lower/higher percentiles [thanks to Geethen for highlighting this]
  • Added examples in comments
  • The documentation has been generated using Sphinx and resides in crepes.readthedocs.io

Fixes

  • Extended type checks to include NumPy floats and integers [thanks to patpizio for pointing this out]
  • Corrected a bug in the assignment of min/max values for Mondrian conformal predictive systems
  • The Jupyter notebook with examples has been updated, changed name to crepes_nb.ipynb and moved to the docs folder
  • Changed the default k to 25 in sigma_knn