linalg.ensemble_lstsq
linalg.ensemble_lstsq(X, Y, n_folds=5, n_reps=20, seed=0, n_jobs=1)
Linear regression with k-fold bagging.
Parameters
| X |
ndarray of shape (n_samples, n_features) |
The design matrix. |
required |
| Y |
ndarray of shape (n_samples, n_targets) |
The target matrix. |
required |
| n_folds |
int |
Number of folds. |
5 |
| n_reps |
int |
Number of repetitions. |
20 |
| seed |
int |
Random seed for the random number generator. |
0 |
Returns
| beta |
ndarray of shape (n_features, n_targets) |
The regression coefficients. |
| Yhat |
ndarray of shape (n_samples, n_targets) |
The predicted values (out-of-bag cross-validation). |
| R2 |
ndarray of shape (n_targets,) |
The R-squared values (cross-validated). |
| r |
ndarray of shape (n_targets,) |
The correlation coefficients between the predicted and actual values. |