将以下Python代码转换为R语言代码# 遗传算法特征选择 sKfold = StratifiedKFold(n_splits=5, shuffle=True, random_state=2020) def fun_(x): '''

以下是转换后的R语言代码： ```R library(caret) library(mlr) # 定义目标函数 fun_ <- function(x) { # 将0/1向量转化为bool类型的特征索引 features <- cols[x == 1] # 提取特征子集 X <- train_1[, features] # 使用特征子集进行交叉验证并计算F1分数 task <- makeClassifTask(data = X, target = "target") lrn <- makeLearner("classif.xgboost", predict.type = "prob") resampling <- makeResampleDesc("CV", iters = 5, stratify = TRUE) measure <- mlr::fmeasure cv <- resample(learner = lrn, task = task, resampling = resampling, measures = measure) val <- mean(cv$measures[[1]]) return(val) } # 定义遗传算法特征选择函数 GV_col <- function(func, len_, num_, it_num = 50, jiaopei = 0.8, bianyi = 0.2) { # 初始化种群 pop <- matrix(sample(0:1, num_ * len_, replace = TRUE), nrow = num_) best_f <- 0 list_best_f <- c() for (i in 1:it_num) { scores <- apply(pop, 1, func) best_fit_ <- max(scores) if (best_fit_ > best_f) { best_f <- best_fit_ best_p <- pop[which.max(scores), ] } list_best_f <- c(list_best_f, best_f) fitness <- scores - min(scores) + 0.01 idx <- sample(num_, num_, replace = TRUE, prob = fitness / sum(fitness)) pop <- pop[idx, ] new_pop <- matrix(0, nrow = num_, ncol = len_) for (j in 1:num_) { father <- pop[j, ] child <- father if (runif(1) < jiaopei) { mother_id <- sample(num_, 1) low_point <- sample(len