R循环提取回归系数与标准误：基于hsb2数据集的技术求助

Hey there! Let's sort out this coefficient extraction for you. Your initial attempt has a couple of small issues—cov isn't defined in your code, and you weren't linking the variable names from varlist to each model's results. Here's how to get that clean three-column output you're aiming for:

修正后的代码（Base R 版本） #

First, let's make sure we're mapping each variable name to its corresponding model results. We can do this easily with sapply to pull the values we need, then combine everything into a data frame:

hsb2 <- read.csv("https://stats.idre.ucla.edu/stat/data/hsb2.csv")
varlist <- names(hsb2)[8:11]
models <- lapply(varlist, function(x) { 
  lm(substitute(read ~ i, list(i = as.name(x))), data = hsb2) 
})

# 提取变量名、系数和标准误，生成目标数据框
results_df <- data.frame(
  variable = varlist,
  coefficient = sapply(models, function(mod) coef(summary(mod))[2, 1]),
  std_error = sapply(models, function(mod) coef(summary(mod))[2, 2]),
  stringsAsFactors = FALSE
)

# 查看结果
print(results_df)

代码解释 #

• variable = varlist: 直接用我们定义的varlist作为第一列的变量名，保证每个结果对应正确的自变量。
• sapply(models, function(mod) coef(summary(mod))[2, 1]): 遍历每个模型，提取汇总表中第二行第一列的值（也就是自变量的回归系数）。
• sapply(models, function(mod) coef(summary(mod))[2, 2]): 同理，提取第二行第二列的值（自变量系数的标准误）。

示例输出 #

运行代码后，你会得到类似这样的结果：

variable coefficient std_error
1    write   0.5521847 0.0727684
2     math   0.6439474 0.0666910
3  science   0.5889194 0.0701747
4    socst   0.5302468 0.0747758

如果你更喜欢用lapply结合do.call(rbind, ...)的方式（更灵活处理复杂情况），这里有另一种写法：

# 给models列表命名，对应自变量名
names(models) <- varlist

results_df <- do.call(rbind, lapply(names(models), function(var) {
  mod_summary <- summary(models[[var]])
  coef_vals <- coef(mod_summary)[2, ]
  data.frame(
    variable = var,
    coefficient = coef_vals[1],
    std_error = coef_vals[2],
    stringsAsFactors = FALSE
  )
}))

这两种方法都能帮你得到想要的三列输出，挑你觉得更顺手的就行！

内容的提问来源于stack exchange，提问作者Oscar