Lasso parameter estimation with path and cross-validationΒΆ
Python source code: lasso_path_with_crossvalidation.py
print __doc__
import numpy as np
################################################################################
# generate some sparse data to play with
n_samples, n_features = 60, 100
np.random.seed(1)
X = np.random.randn(n_samples, n_features)
coef = 3*np.random.randn(n_features)
coef[10:] = 0 # sparsify coef
y = np.dot(X, coef)
# add noise
y += 0.01 * np.random.normal((n_samples,))
# Split data in train set and test set
X_train, y_train = X[:n_samples/2], y[:n_samples/2]
X_test, y_test = X[n_samples/2:], y[n_samples/2:]
################################################################################
# Lasso with path and cross-validation using LassoCV path
from scikits.learn.linear_model import LassoCV
from scikits.learn.cross_val import KFold
cv = KFold(n_samples/2, 5)
lasso_cv = LassoCV()
# fit_params = {'max_iter':100}
y_ = lasso_cv.fit(X_train, y_train, cv=cv, max_iter=100).predict(X_test)
print "Optimal regularization parameter = %s" % lasso_cv.alpha
# Compute explained variance on test data
print "r^2 on test data : %f" % (1 - np.linalg.norm(y_test - y_)**2
/ np.linalg.norm(y_test)**2)