我应该如何将模拟研究中的数据存储在矩阵中？

Question

我目前正在做一个蒙特卡罗模拟，我正在测试哪种模型最适合变量选择。这样做时，我需要将来自每个模拟的信息存储在一个矩阵中，以便以后进行分析。作为 R 的新手，我在执行此任务时遇到了困难。 以下是相关代码：

iter <- 2000

results.matrix <- matrix(ncol = 12, nrow = iter )

for (i in iter) {
  Ylink <- (1 * (age - mean(age)) + (1.5 * (gender == "1")) + 1.2 * (inkomst - mean(inkomst)))
  Yprob <- 1/(1 + exp(-Ylink))
  Y <- rbinom(2000, size = 1, prob = Yprob)
  df <- data.frame(x1, x2, x3, x4, x5, x6, x7, age, gender, inkomst, Y)
  glm0 <- glm(Y ~ x1 + x2 + x3 + x4 + x5 + x6 + x7 + age + gender + inkomst, data = df)
  summary(glm0)
  results.matrix[,1] <- summary(glm0)$coef[, "Pr(>|t|)"]
}

尝试出现以下错误时：“要替换的项目数不是替换长度的倍数”

Answer 1

试试这个：

iter <- 2000

results.matrix <- matrix(ncol = 12, nrow = iter )

for (i in seq_len(iter)) {
  Ylink <- (1 * (age - mean(age)) + (1.5 * (gender == "1")) + 1.2 * (inkomst - mean(inkomst)))
  Yprob <- 1/(1 + exp(-Ylink))
  Y <- rbinom(2000, size = 1, prob = Yprob)
  df <- data.frame(x1, x2, x3, x4, x5, x6, x7, age, gender, inkomst, Y)
  glm0 <- glm(Y ~ x1 + x2 + x3 + x4 + x5 + x6 + x7 + age + gender + inkomst, data = df)
  results.matrix[i,] <- summary(glm0)$coef[, "Pr(>|t|)"]
}