API Reference

This is the class and function reference of mlquantify. Please refer to the full user guide for further details, as the raw specifications of classes and functions may not be enough to give full guidelines on their use. For reference on core concepts, see the Foundations guide.

Object	Description
get_config	Retrieve the current mlquantify configuration. mlquantify
set_config	Set global mlquantify configuration. mlquantify
config_context	Context manager to temporarily change the global mlquantify configuration. mlquantify
BaseQuantifier	Base class for all quantifiers in mlquantify. mlquantify.base
MetaquantifierMixin	Mixin class for meta-quantifiers. mlquantify.base
ProtocolMixin	Mixin class for protocol-based quantifiers. mlquantify.base
AggregationMixin	Mixin class for all aggregative quantifiers. mlquantify.base_aggregative
SoftPredictionMixin	Soft predictions mixin for aggregative quantifiers. mlquantify.base_aggregative
CrispPredictionMixin	Crisp predictions mixin for aggregative quantifiers. mlquantify.base_aggregative
Calibrator	Base class for calibrators. mlquantify.calibration
ClassifierCalibrator	Post-hoc calibration of classifier posteriors by logit scaling. mlquantify.calibration
QuantifierCalibrator	Post-hoc calibration of quantifier prevalence estimates. mlquantify.calibration
BaseComposeQuantifier	Base class for compose-based quantifiers. mlquantify.compose
LinearComposeQuantifier	Compose quantifier for linear representation matching. mlquantify.compose
LikelihoodComposeQuantifier	Compose quantifier based on mixture negative log-likelihood. mlquantify.compose
ComposeQuantifier	Compose quantifier for linear representation matching. mlquantify.compose
BaseConfidenceRegion	Base class for confidence regions of prevalence estimates. mlquantify.confidence
ConfidenceInterval	Bootstrap confidence intervals for each class prevalence. mlquantify.confidence
ConfidenceEllipseSimplex	Confidence ellipse for prevalence estimates in the simplex. mlquantify.confidence
ConfidenceEllipseCLR	Confidence ellipse for prevalence estimates in CLR-transformed space. mlquantify.confidence
construct_confidence_region	Instantiate a confidence region from bootstrap prevalence estimates. mlquantify.confidence
CC	Classify and Count (CC) quantifier. mlquantify.counting
PCC	Probabilistic Classify and Count (PCC) quantifier. mlquantify.counting
ACC	Adjusted Classify and Count (ACC) quantifier. mlquantify.counting
ThresholdAdjustment	Abstract base class for ROC-threshold adjustment quantifiers. mlquantify.counting
TAC	Threshold Adjusted Count (TAC) quantifier. mlquantify.counting
TX	Threshold X (TX) quantifier. mlquantify.counting
TMAX	Threshold MAX (TMAX) quantifier. mlquantify.counting
T50	Threshold 50 (T50) quantifier. mlquantify.counting
MS	Median Sweep (MS) quantifier. mlquantify.counting
MS2	Median Sweep 2 (MS2) quantifier. mlquantify.counting
FM	Friedman Method (FM) quantifier. mlquantify.counting
GACC	Generalized Adjusted Classify and Count (GACC). mlquantify.counting
GPACC	Generalized Probabilistic Adjusted Classify and Count (GPACC). mlquantify.counting
evaluate_thresholds	Evaluate a range of classification thresholds to compute the corresponding True Positive Rate (TPR) and False Positive Rate (FPR) for a binary quantification task. mlquantify.counting
compute_tpr	Compute the True Positive Rate (Recall) for a binary classification task. mlquantify.counting
compute_fpr	Compute the False Positive Rate for a binary classification task. mlquantify.counting
compute_table	Compute the confusion matrix table for a binary classification task. mlquantify.counting
make_quantification	Generate synthetic quantification bags under prior-probability shift. mlquantify.datasets
fetch_mushroom	Mushroom: edible vs. poisonous (binary, all-categorical). mlquantify.datasets
fetch_banknote_authentication	Banknote authentication from wavelet image features (binary). mlquantify.datasets
fetch_haberman_survival	Haberman survival after breast-cancer surgery (binary, hard). mlquantify.datasets
fetch_miniboone	MiniBooNE particle identification: signal vs. background (binary, large). mlquantify.datasets
fetch_digits_optical_penbased	Optical / Pen-based handwritten digits (10-class, easy). mlquantify.datasets
fetch_dry_bean	Dry Bean: seven bean varieties from grain morphology (multiclass). mlquantify.datasets
fetch_covertype	Forest Covertype: 7 cover types from cartographic variables (multiclass, large). mlquantify.datasets
fetch_yeast	Yeast protein localization site (10-class, hard, imbalanced). mlquantify.datasets
fetch_sensorless_drive	Sensorless drive diagnosis from motor current signals (11-class, balanced). mlquantify.datasets
fetch_statlog_shuttle	Statlog (Shuttle): space-shuttle radiator states (multiclass, extreme imbalance). mlquantify.datasets
fetch_wine_quality	Wine Quality: sensory score 3-9 from physicochemistry (ORDINAL). mlquantify.datasets
fetch_online_news_popularity	Online News Popularity: will an article be popular? (binary, temporal). mlquantify.datasets
fetch_pima_diabetes	Pima Indians Diabetes (binary, hard, noisy medical). mlquantify.datasets
fetch_electricity_elec2	Electricity (Elec2): NSW market price up/down stream (binary, drift). mlquantify.datasets
fetch_airlines	Airlines: flight-delay stream (binary, large, temporal). mlquantify.datasets
fetch_newsgroups20	20 Newsgroups: Usenet posts in 20 topics (text, multiclass). mlquantify.datasets
fetch_imdb	IMDB Large Movie Review sentiment (text, binary, balanced). mlquantify.datasets
fetch_multidomain_sentiment	Multi-Domain (Blitzer) Amazon review sentiment (text, covariate shift). mlquantify.datasets
fetch_sentiment140	Sentiment140: 1.6M timestamped tweets (text, binary, temporal). mlquantify.datasets
fetch_rcv1_v2	RCV1-v2: Reuters news topics (text, sparse TF-IDF, multilabel). mlquantify.datasets
fetch_mnist_usps	MNIST -> USPS handwritten digits (image, covariate shift). mlquantify.datasets
fetch_cifar10	CIFAR-10 natural images (image, 10-class, balanced). mlquantify.datasets
fetch_planetoid_cora_citeseer_pubmed	Planetoid citation graphs: Cora / CiteSeer / PubMed (graph nodes). mlquantify.datasets
fetch_sea_concepts	SEA Concepts: synthetic stream with abrupt concept drift (binary). mlquantify.datasets
fetch_lequa2024	LeQua 2024 competition vectors, all tasks via task (text/ordinal). mlquantify.datasets
Bunch	mlquantify.datasets
get_data_home	mlquantify.datasets
fetch_remote	urllib download with local cache + retries (like sklearn). Retries once unverified on TLS errors. mlquantify.datasets
CDE	CDE-Iterate quantifier. mlquantify.likelihood
EMQ	Expectation-Maximization Quantifier (EMQ / SLD). mlquantify.likelihood
MLPE	Maximum Likelihood Prevalence Estimation (MLPE) quantifier. mlquantify.likelihood
BaseLoss	Base class for optimization losses. mlquantify.losses
DistanceLoss	Generic distance-based loss between two probability distributions. mlquantify.losses
LeastSquaresLoss	Squared Euclidean (least-squares) loss. mlquantify.losses
HellingerSurrogateLoss	Optimization surrogate for the squared Hellinger distance. mlquantify.losses
EnergyLoss	Energy-distance loss for distribution matching. mlquantify.losses
NegativeLogLikelihoodLoss	Negative log-likelihood loss for mixture likelihoods. mlquantify.losses
MixtureNegativeLogLikelihoodLoss	Negative log-likelihood for class likelihood mixtures. mlquantify.losses
RegularizedMixtureNLLLoss	Mixture NLL with optional ordinal-smoothness regularization. mlquantify.losses
normalize_distribution	Normalize an array to a valid probability distribution. mlquantify.losses
get_loss	Instantiate a loss object from a string identifier or return a callable. mlquantify.losses
BaseMatchingQuantifier	Base class for distribution matching quantifiers. mlquantify.matching
MatchingHistogramQuantifier	Abstract base class for histogram-based distribution matching. mlquantify.matching
DyS	Distribution y-Similarity (DyS) quantifier. mlquantify.matching
HDy	Hellinger Distance y (HDy) quantifier. mlquantify.matching
HDx	Hellinger Distance x (HDx) quantifier. mlquantify.matching
SORD	Sample Ordinal Distance (SORD) quantifier. mlquantify.matching
MatchingKernelQuantifier	Abstract base class for kernel mean matching quantifiers. mlquantify.matching
MMD_RKHS	Maximum Mean Discrepancy in RKHS (MMD-RKHS) quantifier. mlquantify.matching
KDEyQuantifier	Abstract base class for KDE-based density matching quantifiers. mlquantify.matching
KDEyML	KDEy Maximum Likelihood (KDEy-ML) quantifier. mlquantify.matching
KDEyHD	KDEy Hellinger Distance (KDEy-HD) quantifier. mlquantify.matching
KDEyCS	KDEy Cauchy-Schwarz (KDEy-CS) quantifier. mlquantify.matching
GKDEyML	Generalized KDEy Maximum Likelihood (GKDEyML) quantifier. mlquantify.matching
GHDx	Generalized HDx (GHDx) quantifier. mlquantify.matching
GHDy	Generalized HDy (GHDy) quantifier. mlquantify.matching
SMM	Sample Mean Matching (SMM) quantifier. mlquantify.matching
EDy	Energy Distance y (EDy) quantifier. mlquantify.matching
EDx	Energy Distance x (EDx) quantifier. mlquantify.matching
EnsembleQ	Ensemble Quantifier with prevalence-controlled diversity. mlquantify.meta
QuaDapt	QuaDapt: drift-resilient quantification via parameter adaptation. mlquantify.meta
AggregativeBootstrap	Aggregative Bootstrap quantifier for prevalence confidence regions. mlquantify.meta
AE	Compute the absolute error for each class or a dictionary of errors if input is a dictionary. mlquantify.metrics
SE	Compute the mean squared error between the real and predicted prevalences. mlquantify.metrics
MAE	Compute the mean absolute error between the real and predicted prevalences. mlquantify.metrics
MSE	Mean Squared Error mlquantify.metrics
KLD	Compute the Kullback-Leibler divergence between the real and predicted prevalences. mlquantify.metrics
RAE	Compute the relative absolute error between the real and predicted prevalences. mlquantify.metrics
NAE	Compute the normalized absolute error between the real and predicted prevalences. mlquantify.metrics
NRAE	Compute the normalized relative absolute error between the real and predicted prevalences. mlquantify.metrics
NKLD	Compute the normalized Kullback-Leibler divergence between the real and predicted prevalences. mlquantify.metrics
NMD	Compute the Normalized Match Distance (NMD), also known as Earth Mover’s Distance (EMD), for ordinal quantification evaluation. mlquantify.metrics
RNOD	Compute the Root Normalised Order-aware Divergence (RNOD) for ordinal quantification evaluation. mlquantify.metrics
VSE	Compute the Variance-normalised Squared Error (VSE). mlquantify.metrics
CvM_L1	Compute the L1 version of the Cramér–von Mises statistic (Xiao et al., 2006) between two cumulative distributions, as suggested by Bella et al. (2014). mlquantify.metrics
GridSearchQ	Grid search over quantifier hyperparameters with evaluation protocols. mlquantify.model_selection
BaseProtocol	Abstract base class for evaluation protocols. mlquantify.model_selection
APP	Artificial Prevalence Protocol (APP). mlquantify.model_selection
NPP	Natural Prevalence Protocol (NPP). mlquantify.model_selection
UPP	Uniform Prevalence Protocol (UPP). mlquantify.model_selection
PPP	Personalized Prevalence Protocol (PPP). mlquantify.model_selection
apply_protocol	Evaluate a quantifier across an evaluation protocol. mlquantify.model_selection
binary_quantifier	Decorator to enable binary quantification extensions (One-vs-Rest or One-vs-One). mlquantify.multiclass
BinaryQuantifier	Meta-quantifier enabling One-vs-Rest and One-vs-One strategies. mlquantify.multiclass
MulticlassStrategy	Base class for multiclass decomposition strategies. mlquantify.multiclass
register_strategy	Register a MulticlassStrategy subclass under name. mlquantify.multiclass
get_strategy	Return the registered MulticlassStrategy instance for name. mlquantify.multiclass
available_strategies	Return the sorted names of all registered multiclass strategies. mlquantify.multiclass
PWK	Probabilistic Weighted k-Nearest Neighbour (PWK) quantifier. mlquantify.neighbors
QuaNet	QuaNet: deep neural quantification with an LSTM architecture. mlquantify.neural
BaseRepresentation	Base class for quantification representations. mlquantify.representations
HistogramRepresentation	Histogram-based representation. mlquantify.representations
KDERepresentation	Kernel density estimation representation. mlquantify.representations
DistanceRepresentation	Distance-based representation for quantification. mlquantify.representations
KernelMeanRepresentation	Kernel mean embedding representation. mlquantify.representations
PredictionRepresentation	Representation based on classifier predictions. mlquantify.representations
HardPredictionRepresentation	Hard-prediction representation convenience class. mlquantify.representations
SoftPredictionRepresentation	Soft-prediction representation convenience class. mlquantify.representations
solve_binary	Minimize a scalar objective over the binary prevalence space [0, 1]. mlquantify.solvers
ternary_search	Find the minimum of a unimodal function via ternary search. mlquantify.solvers
solve_simplex	Minimize a function over the probability simplex using SLSQP. mlquantify.solvers
minimize_prevalence	Minimize an objective function over the probability simplex. mlquantify.solvers
minimize_prevalence_blocks	Minimize a loss over multiple representation blocks and aggregate results. mlquantify.solvers
get_prev_from_labels	Get the real prevalence of each class in the target array. mlquantify.utils
normalize_prevalence	Normalize the prevalence of each class to sum to 1. mlquantify.utils
load_quantifier	Load a quantifier from a file. mlquantify.utils
make_prevs	Generate a list of n_dim values uniformly distributed between 0 and 1 that sum exactly to 1. mlquantify.utils
apply_cross_validation	Perform cross-validation and return predictions with true labels for each fold. mlquantify.utils
simplex_uniform_kraemer	Generates n_prev prevalence vectors of n_dim classes uniformly distributed on the simplex, with optional lower and upper bounds. mlquantify.utils
simplex_grid_sampling	Efficiently generates artificial prevalence vectors that sum to 1 and respect min_val ≤ p_i ≤ max_val for all i. mlquantify.utils
simplex_uniform_sampling	Generates uniformly distributed prevalence vectors within the simplex, constrained by min_val ≤ p_i ≤ max_val. mlquantify.utils
get_indexes_with_prevalence	Get indexes for a stratified sample based on the prevalence of each class. mlquantify.utils
DiagonalDisplay	True vs. predicted prevalence diagonal plot. mlquantify.visualization
BiasDisplay	Boxplots of signed prevalence-estimation error. mlquantify.visualization
ErrorByShiftDisplay	Estimation error as a function of prior-probability shift. mlquantify.visualization
PrevalenceDisplay	Bar chart of a single sample’s predicted class prevalence. mlquantify.visualization
ConfidenceRegionDisplay	Confidence region around a single prevalence prediction. mlquantify.visualization