API Reference

This is the class and function reference of cblearn.

cblearn.datasets Datasets

Loaders

datasets.fetch_car_similarity([data_home, ...])	Load the 60-car dataset (most-central triplets).
datasets.fetch_food_similarity([data_home, ...])	Load the Food-100 food similarity dataset (triplets).
datasets.fetch_imagenet_similarity([...])	Load the imagenet similarity dataset (rank 2 from 8).
datasets.fetch_nature_scene_similarity([...])	Load the nature scene similarity dataset (odd-one-out).
datasets.fetch_material_similarity([...])	Load the material similarity dataset (triplets).
datasets.fetch_musician_similarity([...])	Load the MusicSeer musician similarity dataset (triplets).
datasets.fetch_vogue_cover_similarity([...])	Load the vogue cover similarity dataset (odd-one-out).
datasets.fetch_things_similarity([...])	Load the things similarity dataset (odd-one-out).
datasets.fetch_similarity_matrix(name[, ...])	Load human similarity judgements, aggregated to a similarity matrix.

Synthetic Point Generation

datasets.LinearSubspace(subspace_dimension, ...)

Linear Subspace

Simulations

datasets.make_random_triplets(embedding, ...)

Make random triplets with answers for the provided embedding or distances.

Low-level Dataset Utility

datasets.make_all_triplet_indices(n_objects, ...)	Make all triplet indices for a number of objects.
datasets.make_random_triplet_indices(n_objects)	Sample random triplet indices.
datasets.triplet_response(triplets, embedding)	Triplet responses for an embedding.
datasets.noisy_triplet_response(triplets, ...)	Triplet response for an embedding with noise.

cblearn.embedding Embedding

embedding.CKL([n_components, mu, verbose, ...])	Crowd Kernel Learning (CKL) embedding kernel for triplet data.
embedding.FORTE([n_components, verbose, ...])	Fast Ordinal Triplet Embedding (FORTE).
embedding.GNMDS([n_components, lambd, ...])	Generalized Non-metric Multidimensional Scaling (GNMDS).
embedding.SOE([n_components, margin, ...])	Soft Ordinal Embedding (SOE).
embedding.STE([n_components, heavy_tailed, ...])	Stochastic Triplet Embedding algorithm (STE / t-STE).
embedding.TSTE([n_components, verbose, ...])	t-Distributed Stochastic Triplet Embedding (t-STE)
embedding.OENN([n_components, verbose, ...])	Ordinal Embedding Neural Network (OENN).
embedding.MLDS([n_components, random_state, ...])	A maximum-likelihood difference scaling (MLDS) estimator .

Utility

embedding.estimate_dimensionality_cv(...[, ...])	Estimates the dimensionality of the embedding space.
embedding.DimensionEstimationResult(...)	Result object for dimensionality estimation of embeddings.

Wrapper

wrapper.MLDS([n_components, random_state, ...])	A maximum-likelihood difference scaling (MLDS) estimator, wrapping the R implementation.
wrapper.SOE([n_components, n_init, margin, ...])	A soft ordinal embedding estimator, wrapping an R implementation.

cblearn.cluster Cluster

cluster.ComparisonHC(n_clusters)

ComparisonHC.

cblearn.metrics Metrics

metrics.query_accuracy(true_response, ...)	Fraction of violated triplet constraints.
metrics.query_error(true_response, pred_response)	Error measured by 1 - query accuracy.`
metrics.procrustes_distance(true_embedding, ...)	Distance measure between embeddings under optimal transformation.
metrics.query_accuracy_scorer(clf, X, y)	Scorer function for query accuracy, compatible with sklearn's scorer API.

cblearn.preprocessing Preprocessing

preprocessing.query_from_columns(data, ...)	Extract queries with indices from feature columns in a DataFrame.
preprocessing.triplets_from_multiselect(X, ...)	Calculate triplets from n-select or n-rank queries.
preprocessing.triplets_from_oddoneout(X[, y])	Calculates triplets from odd-one-out queries.
preprocessing.triplets_from_mostcentral(X[, y])	Calculates triplets from most-central queries.
preprocessing.SharedColumnEncoder(encoder)	Wrapper to share an encoder across all columns.
preprocessing.MultiColumnLabelEncoder()	Encoder for objects that are a combination of labels in multiple columns.

cblearn.utils Utility

utils.data_format(query[, response])	Extract format of comparison data.
utils.check_format(format, default_query, ...)	Validate comparison format description.
utils.check_query(query[, result_format])	Input validation for queries.
utils.check_query_response(query[, ...])	Input validation for query formats.
utils.check_response(response[, result_format])	Input validation for query responses.
utils.check_size(size, max_objects)	Convert size argument to the number of objects.