Given a dataframe as follows:
city type count
0 bj a 10
1 bj a 23
2 bj b 12
3 bj c 34
4 sh a 17
5 sh b 18
6 sh c 25
7 sh c 13
8 sh a 12
I want to filter rows based on city and type: bj-a, bj-c, sh-b, the expected result will like this:
city type count
0 bj a 10
1 bj a 23
2 bj c 34
3 sh b 18
How can I do that in R? Thanks.
You can do that using subset :
subset(df, city == 'bj' & type %in% c('a', 'c') | city == 'sh' & type == 'b')
# city type count
#0 bj a 10
#1 bj a 23
#3 bj c 34
#5 sh b 18
Or filter in dplyr :
library(dplyr)
df %>%
filter(city == 'bj' & type %in% c('a', 'c') |
city == 'sh' & type == 'b')
For a base R option, you could subset using direct bracket notation:
df[(df$city == "bj" & df$type %in% c("a", "c")) |
(df$city == "sh" & df$type == "b"), ]
city type count
1 bj a 10
2 bj a 23
4 bj c 34
6 sh b 18
Data:
df <- data.frame(city=c("bj", "bj", "bj", "bj", "sh", "sh", "sh", "sh", "sh" ),
type=c("a", "a", "b", "c", "a", "b", "c", "c", "a"),
count=c(10, 23, 12, 34, 17, 18, 25, 13, 12), stringsAsFactors=FALSE)
