lrtree.logreg

implements segment-specific, possibly single-class, logistic regression

class lrtree.logreg.PossiblyOneClassReg(**kwargs)

One class logistic regression (e.g. when a leaf is pure)

Methods

decision_function(X)	Predict confidence scores for samples.
densify()	Convert coefficient matrix to dense array format.
fit(X, y[, weights])	Fit the one class regression: put the label of the single class and the number of features
get_params([deep])	Get parameters for this estimator.
predict(X)	Predict the single class if there's only one class, or predicts
predict_log_proba(X)	Predict logarithm of probability estimates.
predict_proba(X)	Predict the single class with a 1 probability
score(X, y[, sample_weight])	Return the mean accuracy on the given test data and labels.
set_params(**params)	Set the parameters of this estimator.
sparsify()	Convert coefficient matrix to sparse format.

fit(X: np.ndarray | pd.DataFrame, y: np.ndarray, weights: np.ndarray = None)

Fit the one class regression: put the label of the single class and the number of features

predict(X: np.ndarray | pd.DataFrame) → np.ndarray

Predict the single class if there’s only one class, or predicts

Parameters

X (numpy.ndarray or pandas.DataFrame) –

Returns

column of class prediction

Return type

numpy.ndarray

predict_proba(X: np.ndarray | pd.DataFrame)

Predict the single class with a 1 probability

Parameters

X (numpy.ndarray or pandas.DataFrame) –

Returns

two columns, one of which full of ones

Return type

numpy.ndarray

class lrtree.logreg.LogRegSegment(**kwargs)

Logistic Regression for a given segment; fine-tuned PossiblyOneClassReg, itself from sklearn.LogisticRegression, allowing (1) to have only one class and (2) to embed data processing (merging, discretization).

Methods

decision_function(X)	Predict confidence scores for samples.
densify()	Convert coefficient matrix to dense array format.
fit(**kwargs)	Fits the LogRegSegment by preprocessing the data (if need be / asked for) and calling fit on sklearn
get_params([deep])	Get parameters for this estimator.
predict(X)	Predicts the LogRegSegment by preprocessing the data (if need be / asked for) and calling predict on sklearn
predict_log_proba(X)	Predict logarithm of probability estimates.
predict_proba(X)	Predicts the LogRegSegment by preprocessing the data (if need be / asked for) and calling predict_proba on sklearn
score(X, y[, sample_weight])	Return the mean accuracy on the given test data and labels.
set_params(**params)	Set the parameters of this estimator.
sparsify()	Convert coefficient matrix to sparse format.

fit(**kwargs)

Fits the LogRegSegment by preprocessing the data (if need be / asked for) and calling fit on sklearn

predict(X: np.ndarray | pd.DataFrame) → np.ndarray

Predicts the LogRegSegment by preprocessing the data (if need be / asked for) and calling predict on sklearn

predict_proba(X: np.ndarray | pd.DataFrame) → np.ndarray

Predicts the LogRegSegment by preprocessing the data (if need be / asked for) and calling predict_proba on sklearn

Classes

LogRegSegment(**kwargs)	Logistic Regression for a given segment; fine-tuned PossiblyOneClassReg, itself from sklearn.LogisticRegression, allowing (1) to have only one class and (2) to embed data processing (merging, discretization).
PossiblyOneClassReg(**kwargs)	One class logistic regression (e.g.