sklearn.utils

Various utilities to help with development.

Developer guide. See the Utilities for Developers section for further details.

Bunch	Container object exposing keys as attributes.
_safe_indexing	Return rows, items or columns of X using indices.
as_float_array	Convert an array-like to an array of floats.
assert_all_finite	Throw a ValueError if X contains NaN or infinity.
deprecated	Decorator to mark a function or class as deprecated.
estimator_html_repr	Build a HTML representation of an estimator.
gen_batches	Generator to create slices containing batch_size elements from 0 to n.
gen_even_slices	Generator to create n_packs evenly spaced slices going up to n.
indexable	Make arrays indexable for cross-validation.
murmurhash3_32	Compute the 32bit murmurhash3 of key at seed.
resample	Resample arrays or sparse matrices in a consistent way.
safe_mask	Return a mask which is safe to use on X.
safe_sqr	Element wise squaring of array-likes and sparse matrices.
shuffle	Shuffle arrays or sparse matrices in a consistent way.

Input and parameter validation

Functions to validate input and parameters within scikit-learn estimators.

check_X_y	Input validation for standard estimators.
check_array	Input validation on an array, list, sparse matrix or similar.
check_consistent_length	Check that all arrays have consistent first dimensions.
check_random_state	Turn seed into a np.random.RandomState instance.
check_scalar	Validate scalar parameters type and value.
validation.check_is_fitted	Perform is_fitted validation for estimator.
validation.check_memory	Check that memory is joblib.Memory-like.
validation.check_symmetric	Make sure that array is 2D, square and symmetric.
validation.column_or_1d	Ravel column or 1d numpy array, else raises an error.
validation.has_fit_parameter	Check whether the estimator's fit method supports the given parameter.

Meta-estimators

Utilities for meta-estimators.

metaestimators.available_if

An attribute that is available only if check returns a truthy value.

Weight handling based on class labels

Utilities for handling weights based on class labels.

class_weight.compute_class_weight	Estimate class weights for unbalanced datasets.
class_weight.compute_sample_weight	Estimate sample weights by class for unbalanced datasets.

Dealing with multiclass target in classifiers

Utilities to handle multiclass/multioutput target in classifiers.

multiclass.is_multilabel	Check if y is in a multilabel format.
multiclass.type_of_target	Determine the type of data indicated by the target.
multiclass.unique_labels	Extract an ordered array of unique labels.

Optimal mathematical operations

Utilities to perform optimal mathematical operations in scikit-learn.

extmath.density	Compute density of a sparse vector.
extmath.fast_logdet	Compute logarithm of determinant of a square matrix.
extmath.randomized_range_finder	Compute an orthonormal matrix whose range approximates the range of A.
extmath.randomized_svd	Compute a truncated randomized SVD.
extmath.safe_sparse_dot	Dot product that handle the sparse matrix case correctly.
extmath.weighted_mode	Return an array of the weighted modal (most common) value in the passed array.

Working with sparse matrices and arrays

A collection of utilities to work with sparse matrices and arrays.

sparsefuncs.incr_mean_variance_axis	Compute incremental mean and variance along an axis on a CSR or CSC matrix.
sparsefuncs.inplace_column_scale	Inplace column scaling of a CSC/CSR matrix.
sparsefuncs.inplace_csr_column_scale	Inplace column scaling of a CSR matrix.
sparsefuncs.inplace_row_scale	Inplace row scaling of a CSR or CSC matrix.
sparsefuncs.inplace_swap_column	Swap two columns of a CSC/CSR matrix in-place.
sparsefuncs.inplace_swap_row	Swap two rows of a CSC/CSR matrix in-place.
sparsefuncs.mean_variance_axis	Compute mean and variance along an axis on a CSR or CSC matrix.

Utilities to work with sparse matrices and arrays written in Cython.

sparsefuncs_fast.inplace_csr_row_normalize_l1	Normalize inplace the rows of a CSR matrix or array by their L1 norm.
sparsefuncs_fast.inplace_csr_row_normalize_l2	Normalize inplace the rows of a CSR matrix or array by their L2 norm.

Working with graphs

Graph utilities and algorithms.

graph.single_source_shortest_path_length

Return the length of the shortest path from source to all reachable nodes.

Random sampling

Utilities for random sampling.

random.sample_without_replacement

Sample integers without replacement.

Auxiliary functions that operate on arrays

A small collection of auxiliary functions that operate on arrays.

arrayfuncs.min_pos

Find the minimum value of an array over positive values.

Metadata routing

Utilities to route metadata within scikit-learn estimators.

User guide. See the Metadata Routing section for further details.

metadata_routing.MetadataRequest	Contains the metadata request info of a consumer.
metadata_routing.MetadataRouter	Stores and handles metadata routing for a router object.
metadata_routing.MethodMapping	Stores the mapping between caller and callee methods for a router.
metadata_routing.get_routing_for_object	Get a Metadata{Router, Request} instance from the given object.
metadata_routing.process_routing	Validate and route input parameters.

Discovering scikit-learn objects

Utilities to discover scikit-learn objects.

discovery.all_displays	Get a list of all displays from sklearn.
discovery.all_estimators	Get a list of all estimators from sklearn.
discovery.all_functions	Get a list of all functions from sklearn.

API compatibility checkers

Various utilities to check the compatibility of estimators with scikit-learn API.

estimator_checks.check_estimator	Check if estimator adheres to scikit-learn conventions.
estimator_checks.parametrize_with_checks	Pytest specific decorator for parametrizing estimator checks.

Parallel computing

Customizations of joblib and threadpoolctl tools for scikit-learn usage.

parallel.Parallel	Tweak of joblib.Parallel that propagates the scikit-learn configuration.
parallel.delayed	Decorator used to capture the arguments of a function.