R中字符型转数值型出现NAs introduced by coercion警告的技术问询

我有一个包含多个DataFrame的列表，所有数据都以字符型存储，需要调整对应的数据类型。为了避免逐个处理DataFrame，我用循环实现批量操作：先移除无需转为数值型的列（state_ut和_year），将其余字符型列转为数值型；再处理_year列（原格式为YYYY-YY字符型），先转为日期型再提取数值年份，最后合并列。

我的运行代码如下：

library("tidyverse")
temp_num_func <- function(x){
  as.numeric(x, digits = 4) 
}

for(i in 1:length(listofdf)){
  temp_df <- listofdf[[i]]  
  temp_df1 <- select(temp_df, -state_ut, -`_year`) # 移除不需要转数值的列，其余列转数值
  temp_df1 <- temp_df1 %>% mutate_if(is.character, temp_num_func)  
  temp_df2 <- select(temp_df, state_ut, `_year`) 
  temp_df2$`_year` <- as.Date(temp_df2$`_year`, format = "%Y") 
  temp_df2$`_year` <- as.numeric(format(temp_df2$`_year`, "%Y")) 
  listofdf[[i]] <- temp_df2 %>% add_column(temp_df1)
}

但运行后多次收到如下警告：

Warning: NAs introduced by coercionWarning: NAs introduced by coercionWarning: NAs introduced by coercionWarning: NAs introduced by coercionWarning: NAs introduced by coercionWarning: NAs introduced by coercionWarning: NAs introduced by coercionWarning: NAs introduced by coercionWarning: NAs introduced by coercionWarning: NAs introduced by coercionWarning: NAs introduced by coercionWarning: NAs introduced by coercionWarning: NAs introduced by coercionWarning: NAs introduced by coercionWarning: NAs introduced by coercion

相关数据dput输出 #

dput(head(temp_df1[1:4]))的输出： #

structure(list(primary_only = c("88.68", "91.21", "96.95", "89.74",
"100", "100"), primary_with_u_primary = c("95.98", "96.92", "99.03",
"97.37", "100", "100"), primary_with_u_primary_sec_hrsec = c("98.81","99.48", "99.72", "100", "100", "100"), u_primary_only = c("91.39","91.39", "96.32", "0", "100", "0")), row.names = c(NA, 6L), class = "data.frame")

dput(head(listofdf[[1]][1:4]))的输出： #

structure(list(state_ut = c("Andaman & Nicobar Islands", "Andaman & Nicobar Islands", "Andaman & Nicobar Islands", "Andhra Pradesh", "Andhra Pradesh", "Andhra Pradesh"), _year = c("2013-14", "2014-15", "2015-16", "2013-14", "2014-15", "2015-16"), primary_only = c("98.17", "99.55", "100", "86.89", "91.85", "93.89"), primary_with_u_primary = c("98.68", "98.77", "100", "94.53", "96.07","97.02")), row.names = c(NA,6L), class = "data.frame")

问题原因与解决办法 #

1. 警告核心原因：_year列转换逻辑错误 #

你的_year列格式是YYYY-YY（比如2013-14），但你用as.Date(..., format = "%Y")转换时，这个格式仅匹配纯4位年份字符串，导致转换失败生成NA，后续转数值时触发大量警告。

修正_year处理逻辑：直接提取字符串前4位转数值，无需绕日期型转换：

temp_df2$`_year` <- as.numeric(str_sub(temp_df2$`_year`, 1, 4))

2. 简化整体代码（用purrr替代循环更简洁） #

用tidyverse的purrr::map批量处理列表中的DataFrame，代码更简洁易维护：

library(tidyverse)

listofdf <- listofdf %>%
  map(function(df) {
    df %>%
      mutate(
        # 处理_year列：提取前4位转数值
        `_year` = as.numeric(str_sub(`_year`, 1, 4)),
        # 将除state_ut和_year外的字符列统一转数值
        across(-c(state_ut, `_year`), ~as.numeric(.x))
      )
  })

3. 排查潜在的数值转换NA #

如果仍有警告，说明部分列存在非数字字符（比如空字符串、特殊符号），可以用readr::parse_number替代as.numeric，它会自动忽略非数字内容：

across(-c(state_ut, `_year`), ~readr::parse_number(.x))

内容的提问来源于stack exchange，提问作者Rick Bukhariya