mojolearn.QNRegressor#

class mojolearn.QNRegressor(*, loss='squared_error', fit_intercept=True, l1_strength=0.0, l2_strength=0.0, max_iter=1000, tol=0.0001, delta=None, linesearch_max_iter=50, lbfgs_memory=5, warm_start=False, penalty_normalized=True)[source]#

Bases: NumericModeMixin

Linear regression on the squared or absolute loss, solved by a quasi-Newton method.

Minimizes mean_i loss(y_i, x_i @ w + b) + l1_strength * ||w||_1 + (l2_strength / 2) * ||w||^2, where the loss is (z - y)**2 / 2 ('squared_error') or |z - y| ('absolute_error'). With penalty_normalized=True both strengths are divided by the number of samples. The solver is the L-BFGS / OWL-QN solver of LogisticRegression: a nonzero l1_strength selects OWL-QN. The intercept is a solver parameter and is not penalized. This is cuML’s QN solver with loss='l2' or loss='l1'.

X is read as float32; y must be one-dimensional. Out-of-range parameter values raise ValueError.

Parameters:

  • loss ({‘squared_error’, ‘absolute_error’}, default ‘squared_error’) – The data-fit loss; any other value raises ValueError.

  • fit_intercept (bool, default True) – Whether to fit an (unpenalized) intercept.

  • l1_strength (float, default 0.0) – Strength of the l1 penalty, non-negative.

  • l2_strength (float, default 0.0) – Strength of the l2 penalty, non-negative.

  • max_iter (int, default 1000) – Maximum number of solver iterations, a positive integer.

  • tol (float, default 1e-4) – Gradient tolerance, > 0.

  • delta (float or None, default None) – Tolerance on the change of the objective, non-negative; None means tol * 0.01.

  • linesearch_max_iter (int, default 50) – Maximum number of line-search steps per iteration, a positive integer.

  • lbfgs_memory (int, default 5) – Number of past updates kept by L-BFGS, a positive integer.

  • warm_start (bool, default False) – Only False is supported (the solver always starts from zero); True raises NotImplementedError.

  • penalty_normalized (bool, default True) – Whether to divide both penalty strengths by the number of samples.

  • numeric_mode ({‘identical’, ‘fast’} or None, default None) – The numeric mode for this estimator; None uses the process default (see mojolearn.set_numeric_mode). 'fast' is for Apple GPUs.

Variables:

  • coef_ (array of shape (n_features,), float32) – The fitted coefficients.

  • intercept_ (float) – The fitted intercept; 0.0 when fit_intercept is False.

  • n_iter_ (array of shape (1,), int64) – The number of solver iterations run.

  • objective_ (float) – The final value of the minimized objective.

  • retcode_ (int) – The solver’s return code, as on LogisticRegression (0 means converged).

  • n_features_in_ (int) – The number of features seen in fit.

  • input_copied_ (bool) – Whether fit had to copy X to float32 C order.

Examples

>>> import numpy as np, mojolearn
>>> rng = np.random.default_rng(0)
>>> X = rng.random((1000, 8), dtype=np.float32)
>>> y = X @ np.arange(8, dtype=np.float32) + 1.0
>>> model = mojolearn.QNRegressor(loss="absolute_error").fit(X, y)
>>> pred = model.predict(X)
__init__(*, loss='squared_error', fit_intercept=True, l1_strength=0.0, l2_strength=0.0, max_iter=1000, tol=0.0001, delta=None, linesearch_max_iter=50, lbfgs_memory=5, warm_start=False, penalty_normalized=True)[source]#

Methods

fit(X, y[, sample_weight])

Fit the model.

get_params([deep])

numeric_mode_used()

The numeric mode this estimator runs in.

predict(X)

Predict with the fitted model.

score(X, y)

Return the coefficient of determination R^2 of the prediction.

set_params(**params)

Set constructor parameters.

vendor_used()

The GPU backend this estimator runs on.
fit(X, y, sample_weight=None)[source]#

Fit the model.

Parameters:

  • X (array-like of shape (n_samples, n_features)) – Training data, read as float32.

  • y (array-like of shape (n_samples,)) – Target values. A 2-D y or a length mismatch raises ValueError.

  • sample_weight (None) – Must be None; weights raise NotImplementedError.

Returns:

self (QNRegressor) – The fitted estimator.

predict(X)[source]#

Predict with the fitted model.

Parameters:

X (array-like of shape (n_samples, n_features)) – Samples, read as float32.

Returns:

y_pred (array of shape (n_samples,), float32) – X @ coef_ + intercept_.

Raises:

ValueError – If the model is not fitted, or X has a different number of features than the training data.

score(X, y)[source]#

Return the coefficient of determination R^2 of the prediction.

Parameters:

  • X (array-like of shape (n_samples, n_features)) – Test samples.

  • y (array-like of shape (n_samples,)) – True target values.

Returns:

score (float) – 1 - SS_res / SS_tot, computed in float64. A constant y scores 0.0.