How to use regex to subset character format's number?-CodePudding

I have sample dataset like this:

Show in New Window
[1] 84
Show in New Window
structure(list(variable2 = c("ea_level_dataset::constituency", 
"ea_level_dataset::constituency", "ea_level_dataset::constituency", 
"ea_level_dataset::ea_positive_2016", "ea_level_dataset::ea_positive_2016", 
"ea_level_dataset::ea_positive_2016", "ea_level_dataset::ea_positive_2016", 
"ea_level_dataset::ea_positive_2016", "ea_level_dataset::ea_type", 
"ea_level_dataset::ea_type", "ea_level_dataset::households_sprayed_2016", 
"ea_level_dataset::households_sprayed_2016", "ea_level_dataset::households_sprayed_2016", 
"ea_level_dataset::households_sprayed_2016", "ea_level_dataset::households_sprayed_2016", 
"ea_level_dataset::households_sprayed_2016", "ea_level_dataset::households_sprayed_2016", 
"ea_level_dataset::households_sprayed_2016", "ea_level_dataset::households_sprayed_2016", 
"ea_level_dataset::region"), values = c("Kongola", "Linyanti", 
"Sibbinda", "0", "1", "2", "3", "4", "Rural", "Urban", "0", "4", 
"5", "6", "7", "8", "9", "11", "27", "Caprivi"), mappedTerm = c("Kongola", 
"Linyanti", "Sibbinda", "0", "1", "2", "3", "4", "Rural", "Urban", 
"0", "4", "5", "6", "7", "8", "9", "11", "27", "Caprivi"), valueOrder = c(NA_integer_, 
NA_integer_, NA_integer_, NA_integer_, NA_integer_, NA_integer_, 
NA_integer_, NA_integer_, NA_integer_, NA_integer_, NA_integer_, 
NA_integer_, NA_integer_, NA_integer_, NA_integer_, NA_integer_, 
NA_integer_, NA_integer_, NA_integer_, NA_integer_)), class = c("grouped_df", 
"tbl_df", "tbl", "data.frame"), row.names = c(NA, -20L), groups = structure(list(
    variable2 = c("ea_level_dataset::constituency", "ea_level_dataset::ea_positive_2016", 
    "ea_level_dataset::ea_type", "ea_level_dataset::households_sprayed_2016", 
    "ea_level_dataset::region"), .rows = structure(list(1:3, 
        4:8, 9:10, 11:19, 20L), ptype = integer(0), class = c("vctrs_list_of", 
    "vctrs_vctr", "list"))), class = c("tbl_df", "tbl", "data.frame"
), row.names = c(NA, -5L), .drop = TRUE))

Now what i want to do is getting the number value(like 0,1,2,3,4) within each group of variable (those numbers are character format), and remove the words value (like "Rural") in col of mappedTerm.

Could someone help how to do that with regex in R? Thanks~~!

CodePudding user response：

We could use if_all to return only rows having digits by either detecting for one or more digits (\\d ) from the start (^) to end ($) of string,

library(dplyr)
library(stringr)
df1 %>%
   ungroup %>%
   filter(if_all(values:mappedTerm, ~ str_detect(.x, "^\\d $"))) %>% 
   type.convert(as.is = TRUE)

-output

# A tibble: 14 × 4
   variable2                                 values mappedTerm valueOrder
   <chr>                                      <int>      <int> <lgl>     
 1 ea_level_dataset::ea_positive_2016             0          0 NA        
 2 ea_level_dataset::ea_positive_2016             1          1 NA        
 3 ea_level_dataset::ea_positive_2016             2          2 NA        
 4 ea_level_dataset::ea_positive_2016             3          3 NA        
 5 ea_level_dataset::ea_positive_2016             4          4 NA        
 6 ea_level_dataset::households_sprayed_2016      0          0 NA        
 7 ea_level_dataset::households_sprayed_2016      4          4 NA        
 8 ea_level_dataset::households_sprayed_2016      5          5 NA        
 9 ea_level_dataset::households_sprayed_2016      6          6 NA        
10 ea_level_dataset::households_sprayed_2016      7          7 NA        
11 ea_level_dataset::households_sprayed_2016      8          8 NA        
12 ea_level_dataset::households_sprayed_2016      9          9 NA        
13 ea_level_dataset::households_sprayed_2016     11         11 NA        
14 ea_level_dataset::households_sprayed_2016     27         27 NA

Or another option is to force it to numeric with as.numeric and remove the NA elements with complete.cases (will have a warning)

df1 %>%
   ungroup %>%
   mutate(across(values:mappedTerm, as.numeric)) %>%
   filter(if_all(values:mappedTerm, complete.cases))