MLPE#

class mlquantify.likelihood.MLPE(estimator=None, solver='slsqp', tau_0=0.0, tau_1=0.0, cv=None, stratified=True, shuffle=False, random_state=None)[source]#

Maximum Likelihood Prevalence Estimation (MLPE) quantifier.

The trivial quantification baseline. Under the assumption of no shift, the maximum-likelihood estimate of the test prevalence is exactly the observed training prevalence, so MLPE ignores the test set entirely and always returns the training class proportions. It is the reference lower bound that any genuine quantifier should beat; if a method cannot improve on MLPE, it is not exploiting the test data.

Parameters:

estimatorestimator, optional: Kept for API compatibility with the aggregative quantifiers; it does not influence the estimate, since MLPE returns the training prevalence regardless of the test input.
solver, tau_0, tau_1optional: Retained for interface compatibility with the likelihood-composition framework; unused by this trivial baseline.
cvint or None, default=None: Cross-validation folds used when fitting the (unused) estimator.
stratifiedbool, default=True: Whether to stratify CV splits.
shufflebool, default=False: Whether to shuffle data before splitting.
random_stateint or None, default=None: Random seed for reproducibility.

Attributes:

classes_ndarray of shape (n_classes,): Class labels seen during fit.
train_priors_ndarray of shape (n_classes,): Training class prevalences, returned for any test set.

See also

EMQ: Non-trivial maximum-likelihood quantifier (EM re-weighting).
CC: Classify-and-count baseline that does use the test data.

Notes

MLPE makes no use of the test data; it is included only as a sanity-check baseline. Contrast with EMQ, the non-trivial maximum-likelihood quantifier that re-weights the posteriors to the test set.

References

Examples

>>> from mlquantify.likelihood import MLPE
>>> from sklearn.linear_model import LogisticRegression
>>> import numpy as np
>>> rng = np.random.default_rng(0)
>>> X = rng.standard_normal((200, 4))
>>> y = (rng.random(200) < 0.3).astype(int)
>>> q = MLPE(LogisticRegression()).fit(X, y)
>>> q.predict(rng.standard_normal((50, 4)))  # returns the training prevalence
{0: ..., 1: ...}

aggregate(test_representation, train_representation=None, train_labels=None, classes=None)[source]#

Return the training prevalence, ignoring the test representation.

If train_labels is provided its class prevalence is used; otherwise the prevalence observed at fit time (train_priors_) is returned.

fit(X, y, estimator_fitted=False, sample_weight=None, cv_prediction='refit')[source]#

Fit the likelihood compose quantifier.

When representation is provided the fitting is delegated to the parent BaseComposeQuantifier. Otherwise, only the estimator is trained (the likelihood is evaluated directly from posteriors at prediction time).

Parameters:

Xarray-like of shape (n_samples, n_features): Training feature matrix.
yarray-like of shape (n_samples,): Training class labels.
estimator_fittedbool, default=False: Skip fitting the estimator if already fitted.
sample_weightarray-like of shape (n_samples,) or None, default=None: Per-sample weights.
cv_prediction{‘refit’, ‘ensemble’}, default=’refit’: Cross-validation strategy.

Returns:

selfLikelihoodComposeQuantifier: The fitted quantifier.

get_metadata_routing()[source]#

Get metadata routing of this object.

Please check User Guide on how the routing mechanism works.

Returns:

routingMetadataRequest: A MetadataRequest encapsulating routing information.

get_params(deep=True)[source]#

Get parameters for this estimator.

Parameters:

deepbool, default=True: If True, will return the parameters for this estimator and contained subobjects that are estimators.

Returns:

paramsdict: Parameter names mapped to their values.

predict(X)[source]#

Return the training prevalence, ignoring X.

MLPE is the trivial baseline: it estimates the test prevalence as the observed training prevalence, so the test features are not used.

Parameters:

Xarray-like of shape (n_samples, n_features): Test feature matrix (ignored).

Returns:

prevalencesdict or ndarray of shape (n_classes,): The training class prevalences.

save_quantifier(path: str | None = None) → None[source]#: Save the quantifier instance to a file.

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.
False: metadata is not requested and the meta-estimator will not pass it to fit.
None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.
str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters:

cv_predictionstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED: Metadata routing for cv_prediction parameter in fit.
estimator_fittedstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED: Metadata routing for estimator_fitted parameter in fit.
sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED: Metadata routing for sample_weight parameter in fit.

Returns:

selfobject: The updated object.

set_params(**params)[source]#

Set the parameters of this estimator.

The method works on simple estimators as well as on nested objects (such as Pipeline). The latter have parameters of the form <component>__<parameter> so that it’s possible to update each component of a nested object.

Parameters:

**paramsdict: Estimator parameters.

Returns:

selfestimator instance: Estimator instance.