8.14.2.2. sklearn.linear_model.sparse.ElasticNet¶
- class sklearn.linear_model.sparse.ElasticNet(alpha=1.0, rho=0.5, fit_intercept=False, normalize=False, max_iter=1000, tol=0.0001)¶
Linear Model trained with L1 and L2 prior as regularizer
This implementation works on scipy.sparse X and dense coef_.
rho=1 is the lasso penalty. Currently, rho <= 0.01 is not reliable, unless you supply your own sequence of alpha.
Parameters : alpha : float
Constant that multiplies the L1 term. Defaults to 1.0
rho : float
The ElasticNet mixing parameter, with 0 < rho <= 1.
`coef_` : ndarray of shape n_features
The initial coeffients to warm-start the optimization
fit_intercept: bool :
Whether the intercept should be estimated or not. If False, the data is assumed to be already centered.
TODO: fit_intercept=True is not yet implemented
Notes
The parameter rho corresponds to alpha in the glmnet R package while alpha corresponds to the lambda parameter in glmnet.
Methods
decision_function(X) Decision function of the linear model fit(X, y) Fit current model with coordinate descent predict(X) Predict using the linear model score(X, y) Returns the coefficient of determination R^2 of the prediction. set_params(**params) Set the parameters of the estimator. - __init__(alpha=1.0, rho=0.5, fit_intercept=False, normalize=False, max_iter=1000, tol=0.0001)¶
- decision_function(X)¶
Decision function of the linear model
Parameters : X : scipy.sparse matrix of shape [n_samples, n_features] Returns : array, shape = [n_samples] with the predicted real values :
- fit(X, y)¶
Fit current model with coordinate descent
X is expected to be a sparse matrix. For maximum efficiency, use a sparse matrix in CSC format (scipy.sparse.csc_matrix)
- predict(X)¶
Predict using the linear model
Parameters : X : numpy array of shape [n_samples, n_features]
Returns : C : array, shape = [n_samples]
Returns predicted values.
- score(X, y)¶
Returns the coefficient of determination R^2 of the prediction.
The coefficient R^2 is defined as (1 - u/v), where u is the regression sum of squares ((y - y_pred) ** 2).sum() and v is the residual sum of squares ((y_true - y_true.mean()) ** 2).sum(). Best possible score is 1.0, lower values are worse.
Parameters : X : array-like, shape = [n_samples, n_features]
Training set.
y : array-like, shape = [n_samples]
Returns : z : float
- set_params(**params)¶
Set the parameters of the estimator.
The method works on simple estimators as well as on nested objects (such as pipelines). The former have parameters of the form <component>__<parameter> so that it’s possible to update each component of a nested object.
Returns : self :