API Reference

This is the class and function reference of cuML. Please refer to the User Guide for further details, as the raw specifications of classes and functions may not be enough to give full guidelines on their use.

Module Configuration

Output Data Type Configuration

Object	Description	Module
set_global_output_type	Set global output type for cuML estimators.	cuml
using_output_type	Context manager to temporarily set output type.	cuml

Preprocessing, Metrics, and Utilities

Model Selection and Data Splitting

Object	Description	Module
KFold	K-Fold cross-validator.	cuml.model_selection
train_test_split	Split arrays into random train and test subsets.	cuml.model_selection

Feature and Label Encoding (Single-GPU)

Object	Description	Module
LabelBinarizer	Binarize labels in a one-vs-all fashion.	cuml.preprocessing
LabelEncoder	Encode target labels with value between 0 and n_classes-1.	cuml.preprocessing
OneHotEncoder	Encode categorical features as a one-hot numeric array.	cuml.preprocessing
TargetEncoder	Target Encoder for regression and classification targets.	cuml.preprocessing

Feature Scaling and Normalization (Single-GPU)

Object	Description	Module
MaxAbsScaler	Scale each feature by its maximum absolute value.	cuml.preprocessing
MinMaxScaler	Transform features by scaling each feature to a given range.	cuml.preprocessing
Normalizer	Normalize samples individually to unit norm.	cuml.preprocessing
RobustScaler	Scale features using statistics that are robust to outliers.	cuml.preprocessing
StandardScaler	Standardize features by removing the mean and scaling to unit variance.	cuml.preprocessing

Other preprocessing methods (Single-GPU)

Object	Description	Module
ColumnTransformer	Applies transformers to columns of an array or DataFrame.	cuml.compose
PolynomialFeatures	Generate polynomial and interaction features.	cuml.preprocessing
SimpleImputer	Univariate imputer for completing missing values.	cuml.preprocessing

Feature Extraction (Single-GPU)

Object	Description	Module
CountVectorizer	Convert a collection of text documents to a matrix of token counts.	cuml.feature_extraction
HashingVectorizer	Convert a collection of text documents to a matrix of token occurrences.	cuml.feature_extraction
TfidfVectorizer	Convert a collection of raw documents to a matrix of TF-IDF features.	cuml.feature_extraction

Dataset Generation (Single-GPU)

Object	Description	Module
make_blobs	Generate isotropic Gaussian blobs for clustering.	cuml.datasets
make_classification	Generate a random classification problem.	cuml.datasets
make_regression	Generate a random regression problem.	cuml.datasets

Metrics (regression, classification, and distance)

Object	Description	Module
accuracy_score	Accuracy classification score.	cuml.metrics
confusion_matrix	Compute confusion matrix.	cuml.metrics
roc_auc_score	Compute Area Under the ROC Curve.	cuml.metrics

Regression and Classification

Object	Description	Module
BernoulliNB	Naive Bayes classifier for multivariate Bernoulli models.	cuml.naive_bayes
CategoricalNB	Naive Bayes classifier for categorical features.	cuml.naive_bayes
CD	Coordinate Descent solver.	cuml.solvers
ComplementNB	Complement Naive Bayes classifier.	cuml.naive_bayes
ElasticNet	ElasticNet regression.	cuml.linear_model
ForestInference	Forest Inference for fast prediction of tree-based models.	cuml.fil
GaussianNB	Gaussian Naive Bayes.	cuml.naive_bayes
KernelRidge	Kernel Ridge Regression.	cuml.kernel_ridge
KNeighborsClassifier	K-Nearest Neighbors classifier.	cuml.neighbors
KNeighborsRegressor	K-Nearest Neighbors regressor.	cuml.neighbors
Lasso	Lasso regression.	cuml.linear_model
LinearRegression	Ordinary least squares Linear Regression.	cuml.linear_model
LinearSVC	Linear Support Vector Classification.	cuml.svm
LinearSVR	Linear Support Vector Regression.	cuml.svm
LogisticRegression	Logistic Regression classifier.	cuml.linear_model
MBSGDClassifier	Mini Batch SGD Classifier.	cuml.linear_model
MBSGDRegressor	Mini Batch SGD Regressor.	cuml.linear_model
MultinomialNB	Naive Bayes classifier for multinomial models.	cuml.naive_bayes
OneVsOneClassifier	One-vs-one multiclass strategy.	cuml.multiclass
OneVsRestClassifier	One-vs-the-rest multiclass strategy.	cuml.multiclass
QN	Quasi-Newton solver.	cuml.solvers
RandomForestClassifier	Random Forest classifier.	cuml.ensemble
RandomForestRegressor	Random Forest regressor.	cuml.ensemble
Ridge	Ridge regression.	cuml.linear_model
SGD	Stochastic Gradient Descent solver.	cuml.solvers
SVC	C-Support Vector Classification.	cuml.svm
SVR	Epsilon-Support Vector Regression.	cuml.svm

Clustering

Object	Description	Module
AgglomerativeClustering	Agglomerative Clustering.	cuml.cluster
DBSCAN	Density-Based Spatial Clustering of Applications with Noise.	cuml.cluster
HDBSCAN	Hierarchical DBSCAN clustering.	cuml.cluster
KMeans	K-Means clustering.	cuml.cluster
SpectralClustering	Spectral clustering.	cuml.cluster

Dimensionality Reduction and Manifold Learning

Object	Description	Module
GaussianRandomProjection	Reduce dimensionality through Gaussian random projection.	cuml.random_projection
IncrementalPCA	Incremental Principal Component Analysis.	cuml.decomposition
PCA	Principal Component Analysis.	cuml.decomposition
SparseRandomProjection	Reduce dimensionality through sparse random projection.	cuml.random_projection
SpectralEmbedding	Spectral Embedding for non-linear dimensionality reduction.	cuml.manifold
TruncatedSVD	Dimensionality reduction using truncated SVD.	cuml.decomposition
TSNE	t-Distributed Stochastic Neighbor Embedding.	cuml.manifold
UMAP	Uniform Manifold Approximation and Projection.	cuml.manifold

Covariance Estimation

Object	Description	Module
LedoitWolf	Ledoit-Wolf shrinkage covariance estimator.	cuml.covariance

Neighbors

Object	Description	Module
KernelDensity	Kernel Density Estimation.	cuml.neighbors
KNeighborsClassifier	K-Nearest Neighbors classifier.	cuml.neighbors
KNeighborsRegressor	K-Nearest Neighbors regressor.	cuml.neighbors
NearestNeighbors	Unsupervised nearest neighbors.	cuml.neighbors

Time Series

Object	Description	Module
ARIMA	ARIMA time series model.	cuml.tsa
AutoARIMA	Automatic ARIMA model selection.	cuml.tsa
ExponentialSmoothing	Holt-Winters Exponential Smoothing.	cuml.tsa

Model Explainability

Object	Description	Module
KernelExplainer	SHAP Kernel Explainer.	cuml.explainer
PermutationExplainer	SHAP Permutation Explainer.	cuml.explainer

Multi-Node, Multi-GPU Algorithms

Object	Description	Module
CD	Multi-GPU Coordinate Descent solver.	cuml.dask
DBSCAN	Multi-GPU DBSCAN clustering.	cuml.dask
ElasticNet	Multi-GPU ElasticNet regression.	cuml.dask
KMeans	Multi-GPU K-Means clustering.	cuml.dask
KNeighborsClassifier	Multi-GPU K-Nearest Neighbors classifier.	cuml.dask
KNeighborsRegressor	Multi-GPU K-Nearest Neighbors regressor.	cuml.dask
Lasso	Multi-GPU Lasso regression.	cuml.dask
LinearRegression	Multi-GPU Linear Regression.	cuml.dask
MultinomialNB	Multi-GPU Multinomial Naive Bayes.	cuml.dask
NearestNeighbors	Multi-GPU Nearest Neighbors.	cuml.dask
PCA	Multi-GPU Principal Component Analysis.	cuml.dask
RandomForestClassifier	Multi-GPU Random Forest classifier.	cuml.dask
RandomForestRegressor	Multi-GPU Random Forest regressor.	cuml.dask
Ridge	Multi-GPU Ridge Regression.	cuml.dask
TruncatedSVD	Multi-GPU Truncated SVD.	cuml.dask
UMAP	Multi-GPU UMAP.	cuml.dask

cuml.accel

Object	Description	Module
enabled	Check if cuML acceleration is enabled.	cuml.accel
install	Install cuML acceleration hooks.	cuml.accel
is_proxy	Check if an object is a cuML proxy.	cuml.accel
profile	Profile cuML acceleration.	cuml.accel

Experimental

Object	Description	Module
TreeExplainer	SHAP Tree Explainer.	cuml.explainer