jubakit.wrapper package

jubakit.wrapper.classifier module

class jubakit.wrapper.classifier.BaseJubatusClassifier(n_iter=1, shuffle=False, softmax=False, embedded=True, seed=None)[ソース]

ベースクラス: sklearn.base.BaseEstimator, sklearn.base.ClassifierMixin

scikit-learn Wrapper for Jubatus Classifiers.

__init__(n_iter=1, shuffle=False, softmax=False, embedded=True, seed=None)[ソース]

Creates a base class for Jubatus Classifiers.

decision_function(X)[ソース]

Predict confidence scores for samples.

fit(X, y)[ソース]

Fit model.

classmethod get_params(deep=True)[ソース]

Return parameters.

partial_fit(X, y)[ソース]

Partially fit underlying model. If underlying model does not exist, launch a new model.

predict(X)[ソース]

Predict class labels for samples in X.

save(name)[ソース]

Save the classifier model using name.

set_params(**params)[ソース]

Set parameters

stop()[ソース]

Stop the backend process if exists.

class jubakit.wrapper.classifier.LinearClassifier(method=u'AROW', regularization_weight=1.0, softmax=False, n_iter=1, shuffle=False, embedded=True, seed=None)[ソース]

ベースクラス: jubakit.wrapper.classifier.BaseJubatusClassifier

__init__(method=u'AROW', regularization_weight=1.0, softmax=False, n_iter=1, shuffle=False, embedded=True, seed=None)[ソース]
get_params(deep=True)[ソース]
class jubakit.wrapper.classifier.NearestNeighborsClassifier(method=u'euclid_lsh', nearest_neighbor_num=5, local_sensitivity=1.0, hash_num=128, n_iter=1, shuffle=False, softmax=False, embedded=True, seed=None)[ソース]

ベースクラス: jubakit.wrapper.classifier.BaseJubatusClassifier

__init__(method=u'euclid_lsh', nearest_neighbor_num=5, local_sensitivity=1.0, hash_num=128, n_iter=1, shuffle=False, softmax=False, embedded=True, seed=None)[ソース]
get_params(deep=True)[ソース]

jubakit.wrapper.clustering module

class jubakit.wrapper.clustering.BaseJubatusClustering(compressor_method=u'simple', bucket_size=100, compressed_bucket_size=100, bicriteria_base_size=10, bucket_length=2, forgetting_factor=0.0, forgetting_threshold=0.5, seed=0, embedded=True, distance=u'euclidean')[ソース]

ベースクラス: sklearn.base.BaseEstimator, sklearn.base.ClusterMixin

scikit-learn Wrapper for Jubatus Clustering.

__init__(compressor_method=u'simple', bucket_size=100, compressed_bucket_size=100, bicriteria_base_size=10, bucket_length=2, forgetting_factor=0.0, forgetting_threshold=0.5, seed=0, embedded=True, distance=u'euclidean')[ソース]

Creates a base class for Jubatus Clustering

clear()[ソース]
fit_predict(X, y=None)[ソース]

Construct clustering model and Predict the closest cluster each sample in X belongs to.

stop()[ソース]
class jubakit.wrapper.clustering.BaseKFixedClustering(k=2, compressor_method=u'simple', bucket_size=100, compressed_bucket_size=100, bicriteria_base_size=10, bucket_length=2, forgetting_factor=0.0, forgetting_threshold=0.5, seed=0, embedded=True, distance=u'euclidean')[ソース]

ベースクラス: jubakit.wrapper.clustering.BaseJubatusClustering

__init__(k=2, compressor_method=u'simple', bucket_size=100, compressed_bucket_size=100, bicriteria_base_size=10, bucket_length=2, forgetting_factor=0.0, forgetting_threshold=0.5, seed=0, embedded=True, distance=u'euclidean')[ソース]
fit(X, y=None)[ソース]

Construct clustering model.

predict(X)[ソース]

Predict the closest cluster each sample in X belongs to.

class jubakit.wrapper.clustering.DBSCAN(eps=0.2, min_core_point=3, bucket_size=100, compressed_bucket_size=100, bicriteria_base_size=10, bucket_length=2, forgetting_factor=0.0, forgetting_threshold=0.5, seed=0, embedded=True, distance=u'euclidean')[ソース]

ベースクラス: jubakit.wrapper.clustering.BaseJubatusClustering

__init__(eps=0.2, min_core_point=3, bucket_size=100, compressed_bucket_size=100, bicriteria_base_size=10, bucket_length=2, forgetting_factor=0.0, forgetting_threshold=0.5, seed=0, embedded=True, distance=u'euclidean')[ソース]
class jubakit.wrapper.clustering.GMM(k=2, compressor_method=u'simple', bucket_size=100, compressed_bucket_size=100, bicriteria_base_size=10, bucket_length=2, forgetting_factor=0.0, forgetting_threshold=0.5, seed=0, embedded=True, distance=u'euclidean')[ソース]

ベースクラス: jubakit.wrapper.clustering.BaseKFixedClustering

class jubakit.wrapper.clustering.KMeans(k=2, compressor_method=u'simple', bucket_size=100, compressed_bucket_size=100, bicriteria_base_size=10, bucket_length=2, forgetting_factor=0.0, forgetting_threshold=0.5, seed=0, embedded=True, distance=u'euclidean')[ソース]

ベースクラス: jubakit.wrapper.clustering.BaseKFixedClustering

jubakit.wrapper.regression module

class jubakit.wrapper.regression.BaseJubatusRegression(n_iter=1, shuffle=False, embedded=True, seed=None)[ソース]

ベースクラス: sklearn.base.BaseEstimator, sklearn.base.RegressorMixin

scikit-learn Wrapper for Jubatus Regressions.

__init__(n_iter=1, shuffle=False, embedded=True, seed=None)[ソース]

Creates a base class for Jubatus Regressoions.

fit(X, y)[ソース]

Fit model.

classmethod get_params(deep=True)[ソース]

Return parameters.

load(name)[ソース]

Load the regression model using name.

partial_fit(X, y)[ソース]

Partially fit underlying model. If underlying model does not exist, launch a new model.

predict(X)[ソース]

Predict class labels for samples in X.

save(name)[ソース]

Save the regression model using name.

set_params(**params)[ソース]

Set parameters

stop()[ソース]

Stop the backend process if exists.

class jubakit.wrapper.regression.LinearRegression(method=u'AROW', regularization_weight=1.0, sensitivity=1.0, learning_rate=1.0, n_iter=1, shuffle=False, embedded=True, seed=None)[ソース]

ベースクラス: jubakit.wrapper.regression.BaseJubatusRegression

__init__(method=u'AROW', regularization_weight=1.0, sensitivity=1.0, learning_rate=1.0, n_iter=1, shuffle=False, embedded=True, seed=None)[ソース]
get_params(deep=True)[ソース]
class jubakit.wrapper.regression.NearestNeighborsRegression(method=u'euclid_lsh', nearest_neighbor_num=5, hash_num=128, n_iter=1, shuffle=False, embedded=True, seed=None)[ソース]

ベースクラス: jubakit.wrapper.regression.BaseJubatusRegression

__init__(method=u'euclid_lsh', nearest_neighbor_num=5, hash_num=128, n_iter=1, shuffle=False, embedded=True, seed=None)[ソース]
get_params(deep=True)[ソース]