k_fold_cv函数——bartMachine包内函数详解

R bartMachine包下载

R 所有包下载地址 ：
https://cran.r-project.org/web/packages/available_packages_by_name.html

R bartMachine包下载地址 ：
https://cran.r-project.org/web/packages/bartMachine/index.html

一、 函数包内原文

1、k_fold_cv函数底层代码

注：直接运行 k_fold_cv 即可得到

function (X, y, k_folds = 5, folds_vec = NULL, verbose = FALSE, 
    ...) 
{
   
    args = list(...)
    args$serialize = FALSE
    if (class(X) != "data.frame") {
   
        stop("The training data X must be a data frame.")
    }
    if (!(class(y) %in% c("numeric", "integer", "factor"))) {
   
        stop("Your response must be either numeric, an integer or a factor with two levels.\n")
    }
    if (!is.null(folds_vec) & class(folds_vec) != "integer") 
        stop("folds_vec must be an a vector of integers specifying the indexes of each folds.")
    y_levels = levels(y)
    if (class(y) == "numeric" || class(y) == "integer") {
   
        pred_type = "regression"
    }
    else if (class(y) == "factor" & length(y_levels) == 
        2) {
   
        pred_type = "classification"
    }
    n = nrow(X)
    Xpreprocess = pre_process_training_data(X)$data
    p = ncol(Xpreprocess)
    if (is.null(folds_vec)) {
   
        if (k_folds == Inf) {
   
            k_folds = n
        }
        if (k_folds <= 1 || k_folds > n) {
   
            stop("The number of folds must be at least 2 and less than or equal to n, use \"Inf\" for leave one out")
        }
        temp = rnorm(n)
        folds_vec = cut(temp, breaks = quantile(temp, seq(0, 
            1, length.out = k_folds + 1)), include.lowest = T, 
            labels = F)
    }
    else {
   
        k_folds = length(unique(folds_vec))
    }
    if (pred_type == "regression") {
   
        L1_err = 0
        L2_err = 0
        yhat_cv = numeric(n)
    }
    else {
   
        phat_cv = numeric(n)
        yhat_cv = factor(n, levels = y_levels)
        confusion_matrix = matrix(0, nrow = 3, ncol = 3)
        rownames(confusion_matrix) = c(paste("actual", 
            y_levels), "use errors")
        colnames(confusion_matrix) = c(paste("predicted", 
            y_levels), "model errors")
    }
    Xy = data.frame(Xpreprocess, y)
    for (k in 1:k_folds) {
   
        cat(".")
        train_idx = which(folds_vec != k)
        test_idx = setdiff(1:n, train_idx)
        test_data_k = Xy[test_idx, ]
        training_data_k = Xy[train_idx, ]
        bart_machine_cv = do.call(build_bart_machine, c(list(X = training_data_k[,