维度不等时如何从列表元素生成指定格式的数据框？

问题背景 #

现有如下结构的R嵌套列表作为模型输出：

Test <- vector("list" ,2)
names(Test) <- c("Test1", "Test 2")
Test$Test1$factor <- c("Drug1" = 0.036)
Test$Test1$P <- c("Drug1" = 0.066, "P" = 0.69)
Test$Test1$W <- c("Drug1" = 0.084, "W" = 0.0005)

Test$Test2$factor <- c("Drug2" = 0.036)
Test$Test2$P <- c("Drug2" = 0.096, "P" = 0.069)
Test$Test2$W <- c("Drug2" = 0.046, "W" = 0.0105)

需要将每个子列表（Test1、Test 2）转换为指定格式的数据框，示例输出（Test1对应）如下：

data.frame("Resp" = rep("Test1",length(Test$Test1)), 
           Factor  = rep("Drug1",length(Test$Test1)),
           pvalue_factor = c( 0.036, 0.066, 0.084), 
           Vars = c(NA, names(Test$Test1)[-1]),
           pvalue_vars = c(NA, 0.690, 0.0005))

解决方案 #

基础R循环实现 #

无需额外包，直接用循环遍历每个测试项：

# 初始化存储结果的列表
result_list <- list()

# 遍历每个测试名称
for (test_name in names(Test)) {
  current_test <- Test[[test_name]]
  # 提取因子（药物）名称
  factor_name <- names(current_test$factor)
  # 提取所有与因子名称匹配的p值
  p_factor_vals <- unlist(current_test)[names(unlist(current_test)) == factor_name]
  # 提取Vars对应的p值
  vars_p_vals <- unlist(current_test)[names(unlist(current_test)) != factor_name]
  
  # 构建目标数据框
  output_df <- data.frame(
    Resp = rep(test_name, length(current_test)),
    Factor = rep(factor_name, length(current_test)),
    pvalue_factor = p_factor_vals,
    Vars = c(NA, names(current_test)[-1]),
    pvalue_vars = c(NA, vars_p_vals),
    stringsAsFactors = FALSE
  )
  
  result_list[[test_name]] <- output_df
}

# 查看结果
result_list$Test1
result_list$`Test 2`

函数式编程实现（purrr包） #

如果已安装purrr包，可用更简洁的函数式写法：

library(purrr)

# 批量处理所有测试项
result_list <- map(names(Test), function(test_name) {
  current_test <- Test[[test_name]]
  factor_name <- names(current_test$factor)
  p_factor_vals <- unlist(current_test)[names(unlist(current_test)) == factor_name]
  vars_p_vals <- unlist(current_test)[names(unlist(current_test)) != factor_name]
  
  data.frame(
    Resp = rep(test_name, length(current_test)),
    Factor = rep(factor_name, length(current_test)),
    pvalue_factor = p_factor_vals,
    Vars = c(NA, names(current_test)[-1]),
    pvalue_vars = c(NA, vars_p_vals),
    stringsAsFactors = FALSE
  )
}) %>% 
  # 为结果列表命名
  set_names(names(Test))

# 输出结果
print(result_list$Test1)
print(result_list$`Test 2`)

内容的提问来源于stack exchange，提问作者Abbas