Filter "either" "or" with dplyr

5
votes

I have a data.frame that looks like this.

dat1=data.frame(time=c(1,2,1,2,1,2,1,2), team=c("A","A","A","A","B","B","C","C"), name=c("LT","LT","CH","CH","CH","CH","AT","AT"))

time team name
 1    A     LT
 2    A     LT
 1    A     CH
 2    A     CH
 1    B     CH
 2    B     CH
 1    C     AT
 2    C     AT

I would like to say to dplyr, please, group_by "team" and filter the rows of each team that contains the name "LT". If a team does not contain the name "LT", then filter in the name "CH."

I would like my data.frame to look like this.

time team name
 1    A     LT
 2    A     LT
 1    B     CH
 2    B     CH

Any help is highly appreciated

2
rdataframedplyrtidyverse

2 Answers

2
votes

One dplyr option could be:

dat1 %>%
 group_by(team) %>%
 filter(if(any(name == "LT")) name == "LT" else name == "CH")

   time team  name 
  <dbl> <fct> <fct>
1     1 A     LT   
2     2 A     LT   
3     1 B     CH   
4     2 B     CH
2
votes

We can arrange first and then filter based on the first value in 'name' after grouping by 'team'

library(dplyr)
dat1 %>% 
     filter(name %in% c("LT", "CH")) %>% 
     arrange(team, ordered(name, levels = c('LT', 'CH'))) %>% 
     group_by(team) %>%
     filter(name %in% first(name))