Version 0.6.2¶
This is a bug-fix release to resolve some issues regarding the handling the input and the output format of the arrays.
Changelog¶
Allow column vectors to be passed as targets. #673 by Christos Aridas.
Better input/output handling for pandas, numpy and plain lists. #681 by Christos Aridas.
Version 0.6.1¶
This is a bug-fix release to primarily resolve some packaging issues in version 0.6.0. It also includes minor documentation improvements and some bug fixes.
Changelog¶
Bug fixes¶
Fix a bug in
imblearn.ensemble.BalancedRandomForestClassifier
leading to a wrong number of samples used during fitting duemax_samples
and therefore a bad computation of the OOB score. #656 by Guillaume Lemaitre.
Version 0.6.0¶
Changelog¶
Changed models¶
The following models might give some different sampling due to changes in scikit-learn:
The following samplers will give different results due to change linked to the random state internal usage:
Bug fixes¶
imblearn.under_sampling.InstanceHardnessThreshold
now take into account therandom_state
and will give deterministic results. In addition,cross_val_predict
is used to take advantage of the parallelism. #599 by Shihab Shahriar Khan.Fix a bug in
imblearn.ensemble.BalancedRandomForestClassifier
leading to a wrong computation of the OOB score. #656 by Guillaume Lemaitre.
Maintenance¶
Update imports from scikit-learn after that some modules have been privatize. The following import have been changed:
sklearn.ensemble._base._set_random_states
,sklearn.ensemble._forest._parallel_build_trees
,sklearn.metrics._classification._check_targets
,sklearn.metrics._classification._prf_divide
,sklearn.utils.Bunch
,sklearn.utils._safe_indexing
,sklearn.utils._testing.assert_allclose
,sklearn.utils._testing.assert_array_equal
,sklearn.utils._testing.SkipTest
. #617 by Guillaume Lemaitre.Synchronize
imblearn.pipeline
withsklearn.pipeline
. #620 by Guillaume Lemaitre.Synchronize
imblearn.ensemble.BalancedRandomForestClassifier
and add parametersmax_samples
andccp_alpha
. #621 by Guillaume Lemaitre.
Enhancement¶
imblearn.under_sampling.RandomUnderSampling
,imblearn.over_sampling.RandomOverSampling
,imblearn.datasets.make_imbalance
accepts Pandas DataFrame in and will output Pandas DataFrame. Similarly, it will accepts Pandas Series in and will output Pandas Series. #636 by Guillaume Lemaitre.imblearn.FunctionSampler
accepts a parametervalidate
allowing to check or not the inputX
andy
. #637 by Guillaume Lemaitre.imblearn.under_sampling.RandomUnderSampler
,imblearn.over_sampling.RandomOverSampler
can resample when non finite values are present inX
. #643 by Guillaume Lemaitre.All samplers will output a Pandas DataFrame if a Pandas DataFrame was given as an input. #644 by Guillaume Lemaitre.
The samples generation in
imblearn.over_sampling.ADASYN
,imblearn.over_sampling.SMOTE
,imblearn.over_sampling.BorderlineSMOTE
,imblearn.over_sampling.SVMSMOTE
,imblearn.over_sampling.KMeansSMOTE
,imblearn.over_sampling.SMOTENC
is now vectorize with giving an additional speed-up whenX
in sparse. #596 and #649 by Matt Eding.
Deprecation¶
The following classes have been removed after 2 deprecation cycles:
ensemble.BalanceCascade
andensemble.EasyEnsemble
. #617 by Guillaume Lemaitre.The following functions have been removed after 2 deprecation cycles:
utils.check_ratio
. #617 by Guillaume Lemaitre.The parameter
ratio
andreturn_indices
has been removed from all samplers. #617 by Guillaume Lemaitre.The parameters
m_neighbors
,out_step
,kind
,svm_estimator
have been removed from theimblearn.over_sampling.SMOTE
. #617 by Guillaume Lemaitre.