函数式编程问题——map_df & regex

Question

I am trying to combine multiple spreadsheets (about 20) using a functional programming approach. Each spreadsheet contains an individual year of data. They are messy, with columns not named or name of same column changing across the spreadsheets.

I originally did all the cleaning up individually for each spreadsheet but want to learn how to do it with a functional programming to make it more reproducible.

My approach was to build a regex to match all the different names of the specified column, then rename the column using a custom function/regex. I thought I could then use map_dfr to apply this function to all the different spreadsheets to produce a final dataframe to work with.

However I have encountered 2 problems:

1. the regex engine in R seems to have the global parameter on and no way to switch it off. I want to match the the different possibilities in the regex expression in sequence and stop when it finds the first match, not all matches. For example, after I import the spreadsheets sometimes there are mulitple unamed columns which get given names ...1 etc. I only want to match the first instance. I cannot seem to work out if it possible to disable the global parameter, or a cleverer way of writing the regex to stop after the first match. Also is there another, perhaps better, way of approaching this?

2. When I pass my custom function, which seems to work well enough on individual dataframes, I get an error from map_df which I am not quite sure why.

I have produced a minimal reprex below, which I think highlights the issues.

All thoughts greatly received, including alternative approaches to this, as this must be a very common problem people come across. Thanks.

library(tidyverse)

year_1 <- tribble(
  ~`...1`, ~admissions,
  "Hospital 1", 10,
  "Hospital 2", 100,
  "hospital 3", 200
)

year_2 <- tribble(
  ~provider_code, ~`...2`, ~admissions,
  "H1", "Hospital 1", 20,
  "H2", "Hospital 2", 400,
  "H3", "hospital 3", 500
)

year_3 <- tribble(
  ~"Hospital provider code", ~"Commissioning region/Provider", ~admissions,
  "H1", "Hospital 1", 350,
  "H2", "Hospital 2", 350,
  "H3", "hospital 3", 550
)


clean_up_area_column_name <- function(x){
  rename({{x}}, area = matches("\\.{3}[0-9]|commissioning region|hospital provider", ignore.case = TRUE))
  }

clean_up_area_column_name(year_1)
#> # A tibble: 3 × 2
#>   area       admissions
#>   <chr>           <dbl>
#> 1 Hospital 1         10
#> 2 Hospital 2        100
#> 3 hospital 3        200

clean_up_area_column_name(year_2)
#> # A tibble: 3 × 3
#>   provider_code area       admissions
#>   <chr>         <chr>           <dbl>
#> 1 H1            Hospital 1         20
#> 2 H2            Hospital 2        400
#> 3 H3            hospital 3        500

clean_up_area_column_name(year_3)
#> # A tibble: 3 × 3
#>   area1 area2      admissions
#>   <chr> <chr>           <dbl>
#> 1 H1    Hospital 1        350
#> 2 H2    Hospital 2        350
#> 3 H3    hospital 3        550

test_df <- map_dfr(c(year_1, year_2, year_3), clean_up_area_column_name)
#> Error in UseMethod("rename"): no applicable method for 'rename' applied to an object of class "character"

^{Created on 2022-08-08 by the reprex package (v2.0.1)}

Answer 1

Passing multiple data.frames to map requires a list

test_df <- map_dfr(list(year_1, year_2, year_3), clean_up_area_column_name)

# A tibble: 9 x 5
  area       admissions provider_code area1 area2     
  <chr>           <dbl> <chr>         <chr> <chr>     
1 Hospital 1         10 NA            NA    NA        
2 Hospital 2        100 NA            NA    NA        
3 hospital 3        200 NA            NA    NA        
4 Hospital 1         20 H1            NA    NA        
5 Hospital 2        400 H2            NA    NA        
6 hospital 3        500 H3            NA    NA        
7 NA                350 NA            H1    Hospital 1
8 NA                350 NA            H2    Hospital 2
9 NA                550 NA            H3    hospital 3

Answer 2

If you only want to grab the first instances, as you say, then the following tweak to your function should work. Rename any "area1" to "area", then de-select the remaining "area" columns names with trailing digits (area2, area3 etc).

clean_up_area_column_name <- function(x){
rename({{x}}, 
       area = matches("\\.{3}[0-9]|commissioning region|hospital provider")) %>% 
  rename(., area = matches("area1")) %>% 
   select(-any_of(matches("area\\d")))
}

I'm not sure what you expect year_3 to return as it seems your regex is matching the provider_code as well as area :

map_dfr(list(year_1, year_2, year_3), clean_up_area_column_name)

# A tibble: 9 × 3
  area       admissions provider_code
  <chr>           <dbl> <chr>        
1 Hospital 1         10 NA           
2 Hospital 2        100 NA           
3 hospital 3        200 NA           
4 Hospital 1         20 H1           
5 Hospital 2        400 H2           
6 hospital 3        500 H3           
7 H1                350 NA           
8 H2                350 NA           
9 H3                550 NA