TAC#

class mlquantify.adjust_counting.TAC(learner=None, threshold=0.5, strategy='ovr', n_jobs=None)[source]#

Threshold Adjusted Count (TAC) — baseline threshold correction.

This method corrects the bias in class prevalence estimates caused by imperfect classification accuracy, by adjusting the observed positive count using estimates of the classifier’s true positive rate (TPR) and false positive rate (FPR).

It uses a fixed classification threshold and applies the formula:

\[p = \frac{p' - \text{FPR}}{\text{TPR} - \text{FPR}}\]

where $p'$ is the observed positive proportion from CC,

Parameters:

learnerestimator, optional: A supervised learning model with fit and predict_proba methods.
thresholdfloat, default=0.5: Classification threshold in [0, 1] for applying in the CC output.

References

[1]

Forman, G. (2005). “Counting Positives Accurately Despite Inaccurate Classification”, ECML, pp. 564-575.

aggregate(predictions, y_train)[source]#

Aggregate predictions and apply matrix- or rate-based bias correction.

Parameters:

predictionsndarray of shape (n_samples, n_classes): Learner predictions on test data. Can be probabilities (n_samples, n_classes) or class labels (n_samples,).
train_predictionsndarray of shape (n_samples, n_classes): Learner predictions on training data. Can be probabilities (n_samples, n_classes) or class labels (n_samples,).
y_trainndarray of shape (n_samples,): True class labels of the training data.

Returns:

ndarray of shape (n_classes,): Class prevalence estimates.

Examples

>>> from mlquantify.adjust_counting import AC
>>> import numpy as np
>>> q = AC()
>>> predictions = np.random.rand(200)
>>> train_predictions = np.random.rand(200) # generated via cross-validation
>>> y_train = np.random.randint(0, 2, 200)
>>> q.aggregate(predictions, train_predictions, y_train)
{0: 0.51, 1: 0.49}

fit(X, y)[source]#

Fit the quantifier using the provided data and learner.

Parameters:

Xarray-like of shape (n_samples, n_features): Training data.
yarray-like of shape (n_samples,): True labels.
learner_fittedbool, optional: If True, the learner is already fitted, by default False.
cvint, optional: Number of cross-validation folds, by default 5.
stratifiedbool, optional: Whether to stratify the cross-validation, by default True.
random_stateint, optional: Random state for reproducibility, by default None.
shufflebool, optional: Whether to shuffle the data, by default False.

Returns:

selfBaseAdjustCount: Fitted quantifier.

fit_predict(X, y, X_test)[source]#: Fit and predict class prevalences without storing models.

get_best_threshold(thresholds, tprs, fprs)[source]#: Select the best threshold according to the specific method.

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]#: Predict class prevalences for the given data.

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:

cvstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED: Metadata routing for cv parameter in fit.
learner_fittedstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED: Metadata routing for learner_fitted parameter in fit.
random_statestr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED: Metadata routing for random_state parameter in fit.
shufflestr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED: Metadata routing for shuffle parameter in fit.
stratifiedstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED: Metadata routing for stratified 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.

TAC#

This Page