CODEWITHSUNDEEP
rdplyr
Liam Haller
Liam Haller

Reputation: 312

Conditional filtering on spell dataset

In this spell dataset let's say I have user ID, migration number (0,1,2..) and a column for start that represent months from a particular starting date.

sequence_test <- tibble(id = c(1,1,2,2,3,4,4,5),
                         spell_number = c(0,1,0,1,0,0,1,0),
                         start = c(100, 120, 100, 108, 90, 120, 140, 130))

I would like to exclude certain individuals given their start date on their first spell is less than 115, but the indviduals I keep I would like to keep all of their spells even if the start date is before 115.

sequence_test %>% 
  group_by(id) %>% 
  filter(start > 115 (given) spell_number == 1)

I had thought that grouping by IDs since I would like to apply the filtering to all that are the same and then a conditional filter for if start > 115, only for spell_number == 1, but can't find a conditional filtering syntax with dplyr that's not & or |.

Upvotes: 0

Views: 32

Answers (1)

Ma&#235;l
Ma&#235;l

Reputation: 51994

Use any + &:

sequence_test %>% 
  group_by(id) %>% 
  filter(any(start > 115 & spell_number == 1))

     id spell_number start
  <dbl>        <dbl> <dbl>
1     1            0   100
2     1            1   120
3     4            0   120
4     4            1   140

If, for some reason, you don't wanna use &, there is still [:

sequence_test %>% 
  group_by(id) %>% 
  filter(any(start[spell_number == 1] > 115))

Upvotes: 2

Related Questions