Separate cell values with numerical ranges of the form "ab/cd

I'm a newbie at R. The dput of a sample data is given below:

structure(list(fullcode = c("C08J 11/04-11/28", "C22B 7/00-7/04, 19/30, 25/06", 
"B01D 53/00-53/96", "F01N 3/00-3/38", "B01D 45/00-51/00", "A01N 25/00-65/00"
), firstdigits = c("C08J", "C22B", "B01D", "F01N", "B01D", "A01N"
), lastdigits = c("11/04-11/28", "7/00-7/04", "53/00-53/96", 
"3/00-3/38", "45/00-51/00", "25/00-65/00")), row.names = c(NA, 
-6L), class = c("tbl_df", "tbl", "data.frame"))

You'll see that "lastdigits" are comprised of numerical ranges in the form ab/cd-wx/yz. What I want to do is separate these ranges into separate rows, one row per integer. For example, I want to separate the row value 25/00-65/00 into separate rows, each row containing one of 25/00, 25/01, ... 25/99, 26/00, 26/01, ..., 64/98, 64/99, 65/00 in ascending order.

Another example is to separate the row value 53/00-53/96 into separate rows, each containing one of 53/00, 53/01, ..., 53/95, 53/96.

A colleague suggested the following code, but it doesn't work for my case; I believe the below approach assumed that numbers before the \ sign are fixed:

doubled <- grepl("-", range2$lastdigits)
my_data <-
  Map(
    f = function(x, y) {
      out <- range2[x, ]
      if (doubled[[x]]) {
       lastdigits <- gsub(
          "(\\d )/(\\d )-(\\d )/(\\d )",
          "c(\\1, \\2, \\3)", out[["lastdigits"]]
        ) |>
          str2lang() |>
          eval()
        lastdigits[2:3] <- sort(lastdigits[2:3])
        
        out <- data.frame(
          firstdigits = out[["firstdigits"]],
          lastdigits = paste(lastdigits[[1]], seq(from = lastdigits[[2]], to = lastdigits[[3]], by = 1),
                          sep = "/"
          )
        )
      }
      out
    },
    x = seq_len(nrow(range22)),
    y = doubled
  ) |>
  do.call(what = rbind)

Anyone have any ideas as to how I can easily find a solution? Thank you!

CodePudding user response：

We may use

library(dplyr)
library(tidyr)
library(purrr)
df1 %>% 
  separate(lastdigits, into = c('start', 'end'), sep = "-") %>% 
  mutate(across(start:end, ~ as.numeric(chartr("/", ".", .x))), 
     lastdigits = map2(start, end, ~ str_replace(sprintf('%0.2f', 
      seq(.x, .y, by = 0.01)), fixed("."), "/")), start = NULL, end = NULL) %>%
  unnest(lastdigits)

-output

# A tibble: 4,768 × 3
   fullcode         firstdigits lastdigits
   <chr>            <chr>       <chr>     
 1 C08J 11/04-11/28 C08J        11/04     
 2 C08J 11/04-11/28 C08J        11/05     
 3 C08J 11/04-11/28 C08J        11/06     
 4 C08J 11/04-11/28 C08J        11/07     
 5 C08J 11/04-11/28 C08J        11/08     
 6 C08J 11/04-11/28 C08J        11/09     
 7 C08J 11/04-11/28 C08J        11/10     
 8 C08J 11/04-11/28 C08J        11/11     
 9 C08J 11/04-11/28 C08J        11/12     
10 C08J 11/04-11/28 C08J        11/13     
# … with 4,758 more rows

CodePudding user response：

library(dplyr)
library(tidyr)
library(stringr)
df %>%
  mutate(
    temp = str_remove_all(lastdigits, "/"),
    temp = strsplit(temp, split = "-", fixed = TRUE),
    temp = lapply(temp, \(x) as.integer(x[1]):as.integer(x[2]))
  ) %>%
  unnest(temp) %>%
  mutate(final = paste0(substr(temp, 1, 2), "/", substr(temp, 3, 4)))
# # A tibble: 4,768 × 5
#    fullcode         firstdigits lastdigits   temp final
#    <chr>            <chr>       <chr>       <int> <chr>
#  1 C08J 11/04-11/28 C08J        11/04-11/28  1104 11/04
#  2 C08J 11/04-11/28 C08J        11/04-11/28  1105 11/05
#  3 C08J 11/04-11/28 C08J        11/04-11/28  1106 11/06
#  4 C08J 11/04-11/28 C08J        11/04-11/28  1107 11/07
#  5 C08J 11/04-11/28 C08J        11/04-11/28  1108 11/08
#  6 C08J 11/04-11/28 C08J        11/04-11/28  1109 11/09
#  7 C08J 11/04-11/28 C08J        11/04-11/28  1110 11/10
#  8 C08J 11/04-11/28 C08J        11/04-11/28  1111 11/11
#  9 C08J 11/04-11/28 C08J        11/04-11/28  1112 11/12
# 10 C08J 11/04-11/28 C08J        11/04-11/28  1113 11/13
# # … with 4,758 more rows