用 `fread()` 加载 `csv` 文件替换点、逗号和百分比标记

Question

We would like to manage the columns from csv file with originally three character columns class when we used the fread() with the arguments detailed in the R code used section (separator and decimal arguments are specified). R session 版本是4.2.0和data.table版本是1.14.2 。

从csv文件输入数据

col_1,col_2, col_3
/100.432,"30,84 %","4,14"
/3.200,"62,89 %","1,89"
/10.100,"50,00 %","1,62"
/15.570, "40,10 %","3,41"
/900.310, "8,00 %","0,10"

R session 中的输入数据

> dat
# A tibble: 5 × 3
 
  col_1   col_2    col_3
  <chr>  <chr>   <chr>
1 100.432 30,84 % 4,14 
2   3.200 62,89 % 1,89 
3  10.100 50,00 % 1,62 
4  15.570 40,10 % 3,41 
5 900.310  8,00 % 0,10

使用 R 代码

data.table::fread(
  x,
  sep = ',',
  dec = '.',
  na.strings = c('', 'NA')) %>%
as_tibble()

所需的 output数据

> dat
# A tibble: 5 × 3
 
 col_1   col_2  col_3
  <dbl>  <dbl>  <dbl>
1 100438 30.84  4.14 
2   3200 62.89  1.89 
3  10100 50.00  1.62 
4  15570 40.10  3.41 
5 900310  8.00  0.10

问题

我们想获得Desired output data格式。

提前致谢

Answer 1

你可以在 R 中做一些后处理：

RAW = fread('col_1,col_2, col_3
/100.432,"30,84Â %","4,14"
/3.200,"62,89Â %","1,89"
/10.100,"50,00Â %","1,62"
/15.570, "40,10Â %","3,41"
/900.310, "8,00Â %","0,10"')

#       col_1    col_2  col_3
#      <char>   <char> <char>
# 1: /100.432 30,84Â %   4,14
# 2:   /3.200 62,89Â %   1,89
# 3:  /10.100 50,00Â %   1,62
# 4:  /15.570 40,10Â %   3,41
# 5: /900.310  8,00Â %   0,10

library(readr)
RAW[, lapply(.SD, \(x) parse_number(x, locale = locale(decimal_mark = ",")))]
# Or in base R:
RAW[, lapply(.SD, \(x) gsub("[^0-9.]", "", chartr(".,", "_.", x)) |> as.numeric())]

#     col_1 col_2 col_3
#     <num> <num> <num>
# 1: 100432 30.84  4.14
# 2:   3200 62.89  1.89
# 3:  10100 50.00  1.62
# 4:  15570 40.10  3.41
# 5: 900310  8.00  0.10

Answer 2

我改进了@sindri_baldur 提出的后处理。 在dplyr版本1.0.9中，代码应为：

从csv文件输入数据

dat = data.table::fread(
 'col_1,col_2, col_3
 /100.432,"30,84Â %","4,14"
 /3.200,"62,89Â %","1,89"
 /10.100,"50,00Â %","1,62"
 /15.570, "40,10Â %","3,41"
 /900.310, "8,00Â %","0,10"')

R session 中的输入数据

> dat
# A tibble: 5 × 3
 
  col_1   col_2    col_3
  <chr>  <chr>   <chr>
1 100.432 30,84 % 4,14 
2   3.200 62,89 % 1,89 
3  10.100 50,00 % 1,62 
4  15.570 40,10 % 3,41 
5 900.310  8,00 % 0,10

使用 R 代码

> dat <- dat %>%
    as_tibble() %>%
    mutate_at(
      vars(everything()),
      ~ gsub('[^0-9.]', '', chartr('.,', '_.', .))) %>%
    mutate_at(
      vars(everything()),
      ~ as.numeric(as.character(.)))

最终output数据

> dat
# A tibble: 5 × 3

   col_1 col_2 col_3
   <dbl> <dbl> <dbl>
1 100432  30.8  4.14
2   3200  62.9  1.89
3  10100  50    1.62
4  15570  40.1  3.41
5 900310   8    0.1

感谢大家提供解决方案并改进代码。