R通过函数添加前导零时出错

Question

我正在编写代码以更新多个数据集的前导零值。我认为将代码包装在函数中更容易：

convert_scs <- function(df, state,county,street) {

  df$state <- ifelse( nchar(df$state) == 1, paste('0', df$state, sep=''), df$state )

  df$county <- ifelse( nchar(df$county) == 1, paste('00',df$county, sep=''), df$county )
  df$county <- ifelse( nchar(df$county) == 2, paste('0', df$county, sep=''), df$county )

  df$street <- ifelse( (nchar(df$street) == 3) ,  paste('000', df$street, sep=''),  df$street)
  df$street <- ifelse( (nchar(df$street) == 4) ,  paste('00', df$street, sep=''),  df$street)
  df$street <- ifelse( (nchar(df$street) == 5) ,  paste('0', df$street, sep=''),  df$street)

  return(df)

}


convert_scs(predicted20,STATEFP,COUNTYFP, STREETCE)

predicted20 已通过以下方式转换为数据帧：

predicted20 <- as.data.frame(predicted20)

我不断收到以下错误：

$<-.data.frame(*tmp*, "state", value = logical(0)) 中的错误： 替换有0行，数据有1200

数据集中没有缺失。

任何帮助将不胜感激！

预测20的样本

之前的样本数据：

STATEFP(char)  COUNTYFP(char)   STREETCF(char)
1                    1              20100
1                    1              10500
1                    3              11502

需要：

STATEFP(char)  COUNTYFP(char)   STREETCF(char)
01                    001              020100
01                    001              010500
01                    003              011502

Answer 1

我们可以将函数内部的$替换为[[

convert_scs <- function(df, state,county,street) {

  df[[state]] <- ifelse( nchar(df[[state]]) == 1, paste('0', df[[state]], sep=''), df[[state]]
 )

  df[[county]] <- ifelse( nchar(df[[county]]) == 1, paste('00',df[[county]], sep=''), df[[county]] )
 df[[county]] <- ifelse( nchar(df[[county]])  == 2, paste('0',df[[county]] , sep=''),df[[county]]  )

  df[[street]] <- ifelse( (nchar(df[[street]] ) == 3) ,  paste('000',  df[[street]],  sep=''),  df[[street]] )
  df[[street]]  <- ifelse( (nchar( df[[street]] ) == 4) ,  paste('00',  df[[street]] , sep=''),  df[[street]] )
  df[[street]] <- ifelse( (nchar( df[[street]] ) == 5) ,  paste('0',  df[[street]],  sep=''),   df[[street]] )

  df

}

并在函数中将列名作为字符串传递

convert_scs(predicted20,'STATEFP','COUNTYFP', 'STREETCF')
#    STATEFP COUNTYFP STREETCF
#1      01      001   020100
#2      01      001   010500
#3      01      003   011502

---

此外，这可以通过sprintf 轻松完成

library(purrr)
library(dplyr)
predicted20 %>%
   map2_df(.,  c("%02d", "%03d", "%06d"), ~ sprintf(fmt = .y, .x))

或在base R

predicted20[] <-  Map(sprintf, predicted20, fmt = c("%02d", "%03d", "%06d"))

数据

predicted20 <- structure(list(STATEFP = c(1L, 1L, 1L), COUNTYFP = c(1L, 1L, 
3L), STREETCF = c(20100L, 10500L, 11502L)), class = "data.frame", row.names = c(NA, 
-3L))