CODEWITHSUNDEEP
rdplyrfilterconditional-statements
Yvan
Yvan

Reputation: 23

R filter rows such that one column is conditional on two other columns

my data set is as follows:

df <- tibble(id = c("firm a","firm b","firm c","firm d","firm e","firm a","firm e","firm b","firm f","firm g"), 
                     n1 = c(1,1,1,1,1,0,0,0,0,0), 
                     n2 = c(0,0,0,0,0,1,1,1,1,1))

       id        n1    n2
       <chr>  <dbl> <dbl>
     1 firm a     1     0
     2 firm b     1     0
     3 firm c     1     0
     4 firm d     1     0
     5 firm e     1     0
     6 firm a     0     1
     7 firm e     0     1
     8 firm b     0     1
     9 firm f     0     1
    10 firm g     0     1

I want to filter only the rows where the variable id is associated with both a value of 1 in variable n1 and a value of 1 in variable n2.

In this example, the three rows that satisfy this criterion are "firm a", "firm b", and "firm e".

I have tried using the filter function in dplyr, but I could not find how to condition one variable based on values in two other variables.

Upvotes: 1

Views: 899

Answers (2)

Onyambu
Onyambu

Reputation: 79338

df %>%
  group_by(id) %>%
  filter(any(n1 == 1), any(n2 == 1))

# A tibble: 6 x 3
# Groups:   id [3]
  id        n1    n2
  <chr>  <dbl> <dbl>
1 firm a     1     0
2 firm b     1     0
3 firm e     1     0
4 firm a     0     1
5 firm e     0     1
6 firm b     0     1

Upvotes: 0

TarJae
TarJae

Reputation: 79246

After arranging column id we could us lead function and then filter with if_all:

library(dplyr)
df %>% 
  arrange(id) %>% 
  mutate(n2=lead(n2, default = n2[1])) %>% 
  filter(if_all(n1:n2, ~ . == 1))

  id        n1    n2
  <chr>  <dbl> <dbl>
1 firm a     1     1
2 firm b     1     1
3 firm e     1     1

Upvotes: 2

Related Questions