R/LassoPath.R

In belbahrim/tools4uplift: Tools for Uplift Modeling

LassoPath <- function (data, formula) {
    
    # LASSO path for penalized logistic regression.
    #
    # Args:
    #   data: a data frame containing the treatment, the outcome and the predictors.
    #   formula: an object of class "formula" (or one that can be coerced to that class): 
    #            a symbolic description of the model to be fitted.
    # Returns:
    #   A dataframe containing the coefficients values and the number of nonzeros coefficients.
    
    X <- model.matrix(formula, data)
    y <- model.frame(formula, data)[, 1]
    
    get_lambdas <- cv.glmnet(X, 
                             y, 
                             family="binomial",
                             type.measure="auc",
                             nfolds=3)
    
    glmnet.output <- glmnet(X, y, alpha = 1, family = "binomial", 
                            lambda = get_lambdas$lambda, 
                            standardize=TRUE, 
                            intercept = FALSE,
                            type.logistic = "modified.Newton")
    
    
    dimension <- glmnet.output$df
    coeff <- t(as.matrix(glmnet.output$beta))
    lambda <- glmnet.output$lambda
    path <- cbind(lambda, dimension, coeff)
    
    class(path) <- "LassoPath"
    return(path)
}

# END FUN