I would like to create a numeric number to replace characters in R-CodePudding

I have a df1:

ID    Score
XF7    4
XF4    6
XF10   5
DE10   3
XF4    1
DE10   6
DE11   2
R3     8
PPR16  2
DE11   10

I would like to replace the ID's with a number that is unique to the ID. For example XF7 could be "1" while all of the XF4's would be "2".

ID    Score  NumID
XF7    4     1
XF4    6     2
XF10   5     3
DE10   3     4
XF4    1     2
DE10   6     4
DE11   2     5
R3     8     6
PPR16  2     7
DE11   10    5

CodePudding user response：

We may use match

df1$NumID <-  with(df1, match(ID, unique(ID)))

-output

> df1
      ID Score NumID
1    XF7     4     1
2    XF4     6     2
3   XF10     5     3
4   DE10     3     4
5    XF4     1     2
6   DE10     6     4
7   DE11     2     5
8     R3     8     6
9  PPR16     2     7
10  DE11    10     5

CodePudding user response：

Assuming your character IDs are unique, you could convert them to their ASCII encodings.

Define function for converting string to ASCII:

string_to_ASCII <- function(s) {
    paste0(utf8ToInt(s))
}

Replace column of interest:

df1$ID <- lapply(df1$ID, string_to_ASCII)

CodePudding user response：

Another possible solution:

library(dplyr)

df %>% 
  group_by(ID) %>% 
  mutate(NumID = cur_group_id()) %>% 
  ungroup()

#> # A tibble: 10 × 3
#>    ID    Score NumID
#>    <chr> <int> <int>
#>  1 XF7       4     7
#>  2 XF4       6     6
#>  3 XF10      5     5
#>  4 DE10      3     1
#>  5 XF4       1     6
#>  6 DE10      6     1
#>  7 DE11      2     2
#>  8 R3        8     4
#>  9 PPR16     2     3
#> 10 DE11     10     2