Upgrading scikit-learn 0.11 to 0.12
Unknown0.11 → 0.12.1
UnknownUpgrade verdict
DepDelta hasn't looked for this range's notes yet, so breaking changes can't be ruled out.
In plain terms
- DepDelta hasn't finished looking for the notes in this range, so it can't rule out breaking changes yet.
- Their commits are listed with this comparison.
Must act Unknown for releases without notes
Releases without notes may still need changes. Check their commits before upgrading.
API changes
Generated by DepDelta from published Python source. Not mixed with prose notes.
- removed:
export check_build - removed:
export clone - removed:
variable test
- added:
class ClusterMixin(object) - added:
class MetaEstimatorMixin(object) - added:
method ClusterMixin.fit_predict(X, y=…)
- removed:
export check_build - removed:
variable INPLACE_MSG - removed:
variable STANDARD_MSG - removed:
variable dir_content - removed:
variable local_dir - removed:
variable msg
- removed:
function configuration(parent_package=…, top_path=…)
- changed:
class DBSCAN(BaseEstimator) → class DBSCAN(BaseEstimator, ClusterMixin)
- changed:
class Ward(BaseEstimator) → class Ward(BaseEstimator, ClusterMixin) - changed:
constructor Ward(n_clusters=…, memory=…, connectivity=…, copy=…, n_components=…) → constructor Ward(n_clusters=…, memory=…, connectivity=…, copy=…, n_components=…, compute_full_tree=…) - changed:
function ward_tree(X, connectivity=…, n_components=…, copy=…) → function ward_tree(X, connectivity=…, n_components=…, copy=…, n_clusters=…)
- changed:
class MeanShift(BaseEstimator) → class MeanShift(BaseEstimator, ClusterMixin)
- changed:
class SpectralClustering(BaseEstimator) → class SpectralClustering(BaseEstimator, ClusterMixin) - changed:
constructor SpectralClustering(k=…, mode=…, random_state=…, n_init=…) → constructor SpectralClustering(n_clusters=…, mode=…, random_state=…, n_init=…, gamma=…, affinity=…, n_neighbors=…, k=…, precomputed=…) - changed:
function spectral_clustering(affinity, k=…, n_components=…, mode=…, random_state=…, n_init=…) → function spectral_clustering(affinity, n_clusters=…, n_components=…, mode=…, random_state=…, n_init=…, k=…)
- added:
export ledoit_wolf_shrinkage
- changed:
method EmpiricalCovariance.score(X_test, assume_centered=…) → method EmpiricalCovariance.score(X_test) - added:
method EmpiricalCovariance.get_precision()
- changed:
class EllipticEnvelope(OutlierDetectionMixin, MinCovDet) → class EllipticEnvelope(ClassifierMixin, OutlierDetectionMixin, MinCovDet) - changed:
class OutlierDetectionMixin(ClassifierMixin) → class OutlierDetectionMixin(object) - changed:
method OutlierDetectionMixin.decision_function(X, raw_mahalanobis=…) → method OutlierDetectionMixin.decision_function(X, raw_values=…)
- changed:
constructor ShrunkCovariance(store_precision=…, shrinkage=…) → constructor ShrunkCovariance(store_precision=…, assume_centered=…, shrinkage=…) - changed:
function ledoit_wolf(X, assume_centered=…) → function ledoit_wolf(X, assume_centered=…, block_size=…) - changed:
method LedoitWolf.fit(X, assume_centered=…) → method LedoitWolf.fit(X) - changed:
method OAS.fit(X, assume_centered=…) → method OAS.fit(X) - changed:
method ShrunkCovariance.fit(X, assume_centered=…) → method ShrunkCovariance.fit(X) - added:
constructor LedoitWolf(store_precision=…, assume_centered=…, block_size=…) - added:
function ledoit_wolf_shrinkage(X, assume_centered=…, block_size=…)
- removed:
export deprecated - removed:
export load_20newsgroups
- changed:
function make_regression(n_samples=…, n_features=…, n_informative=…, bias=…, effective_rank=…, tail_strength=…, noise=…, shuffle=…, coef=…, random_state=…) → function make_regression(n_samples=…, n_features=…, n_informative=…, n_targets=…, bias=…, effective_rank=…, tail_strength=…, noise=…, shuffle=…, coef=…, random_state=…)
- changed:
function dump_svmlight_file(X, y, f, zero_based=…) → function dump_svmlight_file(X, y, f, zero_based=…, comment=…)
- removed:
export sparse_encode_parallel
- changed:
class FastICA(BaseEstimator) → class FastICA(BaseEstimator, TransformerMixin) - changed:
method FastICA.fit(X) → method FastICA.fit(X, y=…) - changed:
method FastICA.transform(X) → method FastICA.transform(X, y=…) - added:
property FastICA.unmixing_matrix_
- changed:
constructor RandomizedPCA(n_components, copy=…, iterated_power=…, whiten=…, random_state=…) → constructor RandomizedPCA(n_components=…, copy=…, iterated_power=…, whiten=…, random_state=…) - added:
property ProbabilisticPCA.dim
- changed:
constructor MiniBatchSparsePCA(n_components, alpha=…, ridge_alpha=…, n_iter=…, callback=…, chunk_size=…, verbose=…, shuffle=…, n_jobs=…, method=…, random_state=…) → constructor MiniBatchSparsePCA(n_components=…, alpha=…, ridge_alpha=…, n_iter=…, callback=…, chunk_size=…, verbose=…, shuffle=…, n_jobs=…, method=…, random_state=…) - changed:
constructor SparsePCA(n_components, alpha=…, ridge_alpha=…, max_iter=…, tol=…, method=…, n_jobs=…, U_init=…, V_init=…, verbose=…, random_state=…) → constructor SparsePCA(n_components=…, alpha=…, ridge_alpha=…, max_iter=…, tol=…, method=…, n_jobs=…, U_init=…, V_init=…, verbose=…, random_state=…)
- changed:
class BaseEnsemble(BaseEstimator) → class BaseEnsemble(BaseEstimator, MetaEstimatorMixin) - changed:
constructor BaseEnsemble(base_estimator, n_estimators, estimator_params=…) → constructor BaseEnsemble(base_estimator, n_estimators=…, estimator_params=…)
- changed:
constructor GradientBoostingClassifier(loss=…, learn_rate=…, n_estimators=…, subsample=…, min_samples_split=…, min_samples_leaf=…, max_depth=…, init=…, random_state=…) → constructor GradientBoostingClassifier(loss=…, learn_rate=…, n_estimators=…, subsample=…, min_samples_split=…, min_samples_leaf=…, max_depth=…, init=…, random_state=…, max_features=…) - changed:
constructor GradientBoostingRegressor(loss=…, learn_rate=…, n_estimators=…, subsample=…, min_samples_split=…, min_samples_leaf=…, max_depth=…, init=…, random_state=…) → constructor GradientBoostingRegressor(loss=…, learn_rate=…, n_estimators=…, subsample=…, min_samples_split=…, min_samples_leaf=…, max_depth=…, init=…, random_state=…, max_features=…, alpha=…) - added:
class BaseGradientBoosting(BaseEnsemble) - added:
class BinomialDeviance(LossFunction) - added:
class HuberLossFunction(RegressionLossFunction) - added:
class LeastAbsoluteError(RegressionLossFunction) - added:
class LeastSquaresError(RegressionLossFunction) - added:
class LogOddsEstimator(BaseEstimator) - added:
class LossFunction(object) - added:
class MeanEstimator(BaseEstimator) - added:
class MultinomialDeviance(LossFunction) - added:
class PriorProbabilityEstimator(BaseEstimator) - added:
class QuantileEstimator(BaseEstimator) - added:
class QuantileLossFunction(RegressionLossFunction) - added:
class RegressionLossFunction(LossFunction) - added:
constructor BaseGradientBoosting(loss, learn_rate, n_estimators, min_samples_split, min_samples_leaf, max_depth, init, subsample, max_features, random_state, alpha=…) - added:
constructor BinomialDeviance(n_classes) - added:
constructor HuberLossFunction(n_classes, alpha=…) - added:
constructor LossFunction(n_classes) - added:
constructor MultinomialDeviance(n_classes) - added:
constructor QuantileEstimator(alpha=…) - added:
constructor QuantileLossFunction(n_classes, alpha=…) - added:
constructor RegressionLossFunction(n_classes) - added:
field LossFunction.is_multi_class - added:
field MultinomialDeviance.is_multi_class - added:
method BaseGradientBoosting.decision_function(X) - added:
method BaseGradientBoosting.fit(X, y) - added:
method BaseGradientBoosting.fit_stage(i, X, X_argsorted, y, y_pred, sample_mask) - added:
method BaseGradientBoosting.staged_decision_function(X) - added:
method BinomialDeviance.init_estimator() - added:
method BinomialDeviance.negative_gradient(y, pred, **kargs) - added:
method GradientBoostingClassifier.staged_predict(X) - added:
method GradientBoostingClassifier.staged_predict_proba(X) - added:
method HuberLossFunction.init_estimator() - added:
method HuberLossFunction.negative_gradient(y, pred, **kargs) - added:
method LeastAbsoluteError.init_estimator() - added:
method LeastAbsoluteError.negative_gradient(y, pred, **kargs) - added:
method LeastSquaresError.init_estimator() - added:
method LeastSquaresError.negative_gradient(y, pred, **kargs) - added:
method LeastSquaresError.update_terminal_regions(tree, X, y, residual, y_pred, sample_mask, learn_rate=…, k=…) - added:
method LogOddsEstimator.fit(X, y) - added:
method LogOddsEstimator.predict(X) - added:
method LossFunction.init_estimator(X, y) - added:
method LossFunction.negative_gradient(y, y_pred, **kargs) - added:
method LossFunction.update_terminal_regions(tree, X, y, residual, y_pred, sample_mask, learn_rate=…, k=…) - added:
method MeanEstimator.fit(X, y) - added:
method MeanEstimator.predict(X) - added:
method MultinomialDeviance.init_estimator() - added:
method MultinomialDeviance.negative_gradient(y, pred, k=…) - added:
method PriorProbabilityEstimator.fit(X, y) - added:
method PriorProbabilityEstimator.predict(X) - added:
method QuantileEstimator.fit(X, y) - added:
method QuantileEstimator.predict(X) - added:
method QuantileLossFunction.init_estimator() - added:
method QuantileLossFunction.negative_gradient(y, pred, **kargs) - added:
property BaseGradientBoosting.feature_importances_ - added:
variable LOSS_FUNCTIONS
- changed:
constructor PatchExtractor(patch_size, max_patches=…, random_state=…) → constructor PatchExtractor(patch_size=…, max_patches=…, random_state=…)
- changed:
class RFE(BaseEstimator) → class RFE(BaseEstimator, MetaEstimatorMixin) - changed:
class RFECV(RFE) → class RFECV(RFE, MetaEstimatorMixin) - changed:
constructor RFE(estimator, n_features_to_select, step=…) → constructor RFE(estimator, n_features_to_select=…, step=…, estimator_params=…, verbose=…) - changed:
constructor RFECV(estimator, step=…, cv=…, loss_func=…) → constructor RFECV(estimator, step=…, cv=…, loss_func=…, estimator_params=…, verbose=…) - added:
method RFE.decision_function(X) - added:
method RFE.predict_proba(X)
- changed:
constructor GenericUnivariateSelect(score_func, mode=…, param=…) → constructor GenericUnivariateSelect(score_func=…, mode=…, param=…) - changed:
constructor SelectFdr(score_func, alpha=…) → constructor SelectFdr(score_func=…, alpha=…) - changed:
constructor SelectFpr(score_func, alpha=…) → constructor SelectFpr(score_func=…, alpha=…) - changed:
constructor SelectFwe(score_func, alpha=…) → constructor SelectFwe(score_func=…, alpha=…) - changed:
constructor SelectKBest(score_func, k=…) → constructor SelectKBest(score_func=…, k=…) - changed:
constructor SelectPercentile(score_func, percentile=…) → constructor SelectPercentile(score_func=…, percentile=…)
- removed:
variable EPS - removed:
variable NEGINF - removed:
variable ZEROLOGPROB - changed:
constructor GMMHMM(n_components=…, n_mix=…, startprob=…, transmat=…, startprob_prior=…, transmat_prior=…, algorithm=…, gmms=…, covariance_type=…, covars_prior=…, random_state=…) → constructor GMMHMM(n_components=…, n_mix=…, startprob=…, transmat=…, startprob_prior=…, transmat_prior=…, algorithm=…, gmms=…, covariance_type=…, covars_prior=…, random_state=…, n_iter=…, thresh=…, params=…, init_params=…) - changed:
constructor GaussianHMM(n_components=…, covariance_type=…, startprob=…, transmat=…, startprob_prior=…, transmat_prior=…, algorithm=…, means_prior=…, means_weight=…, covars_prior=…, covars_weight=…, random_state=…) → constructor GaussianHMM(n_components=…, covariance_type=…, startprob=…, transmat=…, startprob_prior=…, transmat_prior=…, algorithm=…, means_prior=…, means_weight=…, covars_prior=…, covars_weight=…, random_state=…, n_iter=…, thresh=…, params=…, init_params=…) - changed:
constructor MultinomialHMM(n_components=…, startprob=…, transmat=…, startprob_prior=…, transmat_prior=…, algorithm=…, random_state=…) → constructor MultinomialHMM(n_components=…, startprob=…, transmat=…, startprob_prior=…, transmat_prior=…, algorithm=…, random_state=…, n_iter=…, thresh=…, params=…, init_params=…)
- removed:
export LARS - removed:
export LassoLARS - added:
export MultiTaskElasticNet - added:
export MultiTaskLasso
- removed:
class BaseSGD(BaseEstimator) - removed:
constructor BaseSGD(loss, penalty=…, alpha=…, rho=…, fit_intercept=…, n_iter=…, shuffle=…, verbose=…, seed=…, learning_rate=…, eta0=…, power_t=…, warm_start=…) - removed:
method BaseSGD.fit(X, y) - removed:
method BaseSGD.predict(X) - changed:
class LinearModel(BaseEstimator, RegressorMixin) → class LinearModel(BaseEstimator) - changed:
class LinearRegression(LinearModel) → class LinearRegression(LinearModel, RegressorMixin) - added:
function sparse_center_data(X, y, fit_intercept, normalize=…) - added:
method LinearModel.fit(X, y)
- changed:
class OrthogonalMatchingPursuit(LinearModel) → class OrthogonalMatchingPursuit(LinearModel, RegressorMixin)
- added:
constructor BaseRandomizedLinearModel()
- removed:
constructor RidgeCV(alphas=…, fit_intercept=…, normalize=…, score_func=…, loss_func=…, cv=…, gcv_mode=…) - removed:
method Ridge.fit(X, y, sample_weight=…, solver=…) - removed:
method RidgeCV.fit(X, y, sample_weight=…) - removed:
method RidgeClassifier.decision_function(X) - removed:
method RidgeClassifierCV.decision_function(X) - changed:
class Ridge(LinearModel) → class Ridge(_BaseRidge, RegressorMixin) - changed:
class RidgeCV(LinearModel) → class RidgeCV(_BaseRidgeCV, RegressorMixin) - changed:
class RidgeClassifier(Ridge) → class RidgeClassifier(_BaseRidge, ClassifierMixin) - changed:
class RidgeClassifierCV(RidgeCV) → class RidgeClassifierCV(_BaseRidgeCV, ClassifierMixin)
- removed:
constructor ElasticNet(alpha=…, rho=…, fit_intercept=…, normalize=…, max_iter=…, tol=…) - removed:
constructor Lasso(alpha=…, fit_intercept=…, normalize=…, max_iter=…, tol=…) - removed:
method ElasticNet.decision_function(X) - removed:
method ElasticNet.fit(X, y) - changed:
class ElasticNet(LinearModel) → class ElasticNet(ElasticNet) - changed:
class Lasso(ElasticNet) → class Lasso(Lasso)
- removed:
function configuration(parent_package=…, top_path=…)
- changed:
constructor SGDClassifier(loss=…, penalty=…, alpha=…, rho=…, fit_intercept=…, n_iter=…, shuffle=…, verbose=…, n_jobs=…, seed=…, learning_rate=…, eta0=…, power_t=…, class_weight=…, warm_start=…) → constructor SGDClassifier(loss=…, penalty=…, alpha=…, rho=…, fit_intercept=…, n_iter=…, shuffle=…, verbose=…, epsilon=…, n_jobs=…, seed=…, learning_rate=…, eta0=…, power_t=…, class_weight=…, warm_start=…) - changed:
constructor SGDRegressor(loss=…, penalty=…, alpha=…, rho=…, fit_intercept=…, n_iter=…, shuffle=…, verbose=…, p=…, seed=…, learning_rate=…, eta0=…, power_t=…, warm_start=…) → constructor SGDRegressor(loss=…, penalty=…, alpha=…, rho=…, fit_intercept=…, n_iter=…, shuffle=…, verbose=…, epsilon=…, p=…, seed=…, learning_rate=…, eta0=…, power_t=…, warm_start=…) - added:
class BaseSGD(BaseEstimator) - added:
constructor BaseSGD(loss, penalty=…, alpha=…, rho=…, fit_intercept=…, n_iter=…, shuffle=…, verbose=…, epsilon=…, seed=…, learning_rate=…, eta0=…, power_t=…, warm_start=…) - added:
field SGDClassifier.loss_functions - added:
field SGDRegressor.loss_functions - added:
method BaseSGD.fit(X, y) - added:
method BaseSGD.predict(X) - added:
method BaseSGD.set_params(*args, **kwargs) - added:
method SGDClassifier.predict_log_proba(X) - added:
variable DEFAULT_EPSILON - added:
variable LEARNING_RATE_TYPES - added:
variable PENALTY_TYPES - added:
variable SPARSE_INTERCEPT_DECAY
- added:
export MDS
- changed:
class Isomap(BaseEstimator) → class Isomap(BaseEstimator, TransformerMixin)
- changed:
class LocallyLinearEmbedding(BaseEstimator) → class LocallyLinearEmbedding(BaseEstimator, TransformerMixin)
- added:
export auc_score - added:
export average_precision_score - added:
export normalized_mutual_info_score
- added:
export entropy - added:
export normalized_mutual_info_score
- added:
function normalized_mutual_info_score(labels_true, labels_pred)
- added:
function auc_score(y_true, y_score) - added:
function average_precision_score(y_true, y_score)
- removed:
function euclidian_distances(*args, **kwargs)
- added:
export _validate_covars
- changed:
class OneVsOneClassifier(BaseEstimator, ClassifierMixin) → class OneVsOneClassifier(BaseEstimator, ClassifierMixin, MetaEstimatorMixin) - changed:
class OneVsRestClassifier(BaseEstimator, ClassifierMixin) → class OneVsRestClassifier(BaseEstimator, ClassifierMixin, MetaEstimatorMixin) - changed:
class OutputCodeClassifier(BaseEstimator, ClassifierMixin) → class OutputCodeClassifier(BaseEstimator, ClassifierMixin, MetaEstimatorMixin) - added:
property OneVsRestClassifier.classes_
- removed:
class BaseDiscreteNB(BaseNB) - removed:
class BaseNB(BaseEstimator, ClassifierMixin) - removed:
field BaseDiscreteNB.coef_ - removed:
field BaseDiscreteNB.intercept_ - removed:
method BaseDiscreteNB.fit(X, y, sample_weight=…, class_prior=…) - removed:
method BaseNB.predict(X) - removed:
method BaseNB.predict_log_proba(X) - removed:
method BaseNB.predict_proba(X) - removed:
property GaussianNB.class_prior - removed:
property GaussianNB.sigma - removed:
property GaussianNB.theta
- added:
export NearestCentroid
- added:
constructor NeighborsBase()
- added:
method KNeighborsClassifier.predict_proba(X)
- changed:
class PLSSVD(BaseEstimator) → class PLSSVD(BaseEstimator, TransformerMixin) - added:
method PLSSVD.fit_transform(X, y=…, **fit_params)
- changed:
method KernelCenterer.fit(K) → method KernelCenterer.fit(K, y=…) - changed:
method KernelCenterer.transform(K, copy=…) → method KernelCenterer.transform(K, y=…, copy=…) - added:
class LabelEncoder(BaseEstimator, TransformerMixin) - added:
method LabelEncoder.fit(y) - added:
method LabelEncoder.fit_transform(y) - added:
method LabelEncoder.inverse_transform(y) - added:
method LabelEncoder.transform(y)
- added:
property QDA.classes
- added:
export libsvm_sparse
- changed:
class NuSVC(BaseLibSVM, ClassifierMixin) → class NuSVC(BaseSVC) - changed:
class SVC(BaseLibSVM, ClassifierMixin) → class SVC(BaseSVC)
- changed:
class NuSVC(SparseBaseLibSVM, ClassifierMixin) → class NuSVC(SparseBaseLibSVM, BaseSVC) - changed:
class SVC(SparseBaseLibSVM, ClassifierMixin) → class SVC(SparseBaseLibSVM, BaseSVC)
- added:
function safe_sqr(X, copy=…)
- removed:
function fast_svd(M, k, p=…, n_iterations=…, transpose=…, random_state=…) - added:
function pinvh(a, cond=…, rcond=…, lower=…)
- removed:
function dot_out(a, b, out=…) - removed:
variable dot_out - added:
function safe_copy(X) - added:
variable safe_copy
- changed:
function array2d(X, dtype=…, order=…) → function array2d(X, dtype=…, order=…, copy=…) - changed:
function atleast2d_or_csr(X, dtype=…, order=…) → function atleast2d_or_csr(X, dtype=…, order=…, copy=…) - added:
function atleast2d_or_csc(X, dtype=…, order=…, copy=…)