Filtering out spouses from respondent-spouse groups in survey data

here is a small dataframe that is a simplification of what I am working with:

data.frame(resp = seq(1, 10), spouse = c(2, 1, 5, NA, 3, 3, NA, 10, NA, 8), outcome = seq(11, 20, 1)) -> df
df <- df[sample(1:nrow(df)), ]

Each respondent is identified by a unique identifier in the resp column. However, some respondents are spouses of other respondents, and so I must remove them to prevent reverse causality later on.
This would be what I want to get in the end (not in o
ideal_df <- df %>% filter(resp %in% c(1, 3, 4, 7, 8, 9))
I cannot just use filter because that would remove all spouses in the group and leave me with the singles and that third wheel. Like so:

df %>% filter(!(resp %in% spouse))

I can't group by spouse and respondent groups either since spouses and respondents have different identifiers.
For the full dataset, I am using the RAND HRS dataset distributed by University of Michigan.
I could not just upload the entire set since I have no idea where the spouses are located, and so all I can do is subset by row and that might just leave all the spouses out.
Please let me know what I can do to either improve this question/where I can go/what I should do. Thank you very much for your help.

If you just want to keep one of each pair and it does not matter which one you can simply keep one of them by adding an additional condition to your filter:

df %>% filter(!((resp %in% spouse) & (spouse > resp)))

Output would be all the singles plus one individual from each of the couples:

# OUTPUT
  resp spouse outcome
1    9     NA      19
2    6      3      16
3    1      2      11
4    8     10      18
5    4     NA      14
6    3      5      13
7    7     NA      17

you can add a couple id and keep the distinct ones with dplyr distinct.

with the ifelse statement below you have:

• The same couple id for resp and spouse whenever is the rep id is first

• and id_single if spouse is missing.

library(dplyr)

df %>%
  mutate(couple=ifelse(resp>spouse,
                       paste0(resp,"_",spouse),
                       paste0(spouse,"_",resp)))%>%
  mutate(couple=ifelse(is.na(couple),paste0(resp,"_single"),couple))%>%
  distinct(couple,.keep_all = TRUE)
#>   resp spouse outcome   couple
#> 1    1      2      11      2_1
#> 2    3      5      13      5_3
#> 3    4     NA      14 4_single
#> 4    6      3      16      6_3
#> 5    7     NA      17 7_single
#> 6    8     10      18     10_8
#> 7    9     NA      19 9_single

^{Created on 2023-04-13 with reprex v2.0.2}