Best way to exclude any of multiple characters in a string column when filtering a dataframe in R

I have a dataframe with string and number data that I need to filter.

library(dplyr)
df <- data.frame (Name  = c("101a,102a", "101b,102b,103b", "103c", "102d,103d", "101,103"),
                  Value = c("2", "3", "4", "5", "6")
                  )

The supplementary characters in Name can only be a, b, c or d. I want to find the best way to filter away all data where other characters than the one provided in the filter occur, while keeping data where Name does not contain characters. When filtering with "a", I wish to remove all data that contains "b", "c" or "d", keeping the first and last data:

Name                Value
"101a,102a"         "2"
"101,102,103"       "6"

I can probably do this with if elses

If (char=="a") {
	df <- filter(df, (!grepl("b", Name) | !grepl("c", Name) | !grepl("d", Name))
} else if (char=="b") {
	df <- filter(df, (!grepl("a", Name) | !grepl("c", Name) | !grepl("d", Name))
} else if (char=="c") {
	df <- filter(df, (!grepl("a", Name) | !grepl("b", Name) | !grepl("d", Name))
} else if (char=="d") {
	df <- filter(df, (!grepl("a", Name) | !grepl("b", Name) | !grepl("c", Name))
}

But I was hoping someone could help me to something more efficient and shorter code. I'm looking for a code that essentially does this:

"remove char from 'a,b,c,d' and filter out all data where Name does not contain any the remaining chars".

I tried:

abcd <- c("a", "b", "c", "d")
df <- filter(df, !Name %in% abcd[!abcd==char])

but %in% seems to use match which requires perfect match, so I tried

df <- filter(!grepl(paste(abcd[!abcd==char], collapse="|"),Name))

but I can't get the right syntax. I think I need some help creating the

(!grepl("a", Name) | !grepl("b", Name) | !grepl("c", Name))

part on the fly.

If you only ever do one letter as a time you could use a function like this.

library(dplyr)
df <- data.frame (Name  = c("101a,102a", "101b,102b,103b", "103c", "102d,103d", "101,103"),
                  Value = c("2", "3", "4", "5", "6")
)


keep <- function(df, keep = c("a", "b", "c", "d")){
     df[grepl(keep, df$Name,  fixed = TRUE),]
}

> keep(df, "a")
       Name Value
1 101a,102a     2

> keep(df, "b")
            Name Value
2 101b,102b,103b     3

Use paste to create a regex not including the character you want. Then filter negating the result of grepl.

suppressPackageStartupMessages(
  library(dplyr)
)

df <- data.frame(Name  = c("101a,102a", "101b,102b,103b", "103c", "102d,103d", "101,103"),
                 Value = c("2", "3", "4", "5", "6"))

abcd <- c("a", "b", "c", "d")

char <- "a"

discard <- paste(abcd[abcd != char], collapse = "|")
filter(df, !grepl(discard, Name))
#>        Name Value
#> 1 101a,102a     2
#> 2   101,103     6

^{Created on 2023-05-28 with reprex v2.0.2}

A base R way is the following.

char <- "a"

discard <- paste(abcd[abcd != char], collapse = "|")
df[grep(discard, df$Name, invert = TRUE), ]
#>        Name Value
#> 1 101a,102a     2
#> 5   101,103     6

^{Created on 2023-05-28 with reprex v2.0.2}

Here is solution with bringing the data in long format with separate_rows():

library(tidyverse)

df %>% 
  separate_rows(Name) %>% 
  mutate(x = str_extract(Name, "[A-Za-z]"),
         Name = parse_number(Name)) %>% 
  filter(x == "a" | is.na(x)) %>% 
  mutate(Name = ifelse(!is.na(x), paste0(Name, x), Name)) %>% 
  summarise(Name = toString(Name), .by=Value)

  Value Name      
  <chr> <chr>     
1 2     101a, 102a
2 6     101, 103