CODEWITHSUNDEEP
rfilterdplyr
phoebe
phoebe

Reputation: 9

filter based on negative statement in R dplyr

i'm trying to filter based on a negative statement joined by & operator. the positive statement gives me the correct stuff i do not want, but somehow when i add in ! to indicate the negative statement (e.g., !(logic1 & logic2)), it's not returning correctly.

here are the specifics, i want to exclude rows with ID that is in bad AND tagged as "no good" in the column tag. 

bad <- c(101, 103, 107, 110)

the following positive statement returns the correct rows i do not want:

df %>%
 filter(ID %in% bad & tag == "no good")

so then i added ! to indicate the negative:

df %>% filter (!(ID %in% bad & tag == "no good"))

somehow, this kicked out the folks with IDs matching bad regardless of whether tag had "no good". iow, folks with IDs matching bad were kicked out even when tag was blank (NA_character_).

what am i doing wrong here?

thank you!

Upvotes: 0

Views: 5008

Answers (2)

akrun
akrun

Reputation: 887088

We can modify to |

library(dplyr)
df %>% 
     filter (!(ID %in% bad | tag == "no good"))
#    ID  tag
#1 105 good
#2 109 good

data

bad <- c(101, 103, 107, 110)
df  <- data.frame(ID=c(101:110),tag = rep(c("good","no good"),5))

Upvotes: 1

sachin2014
sachin2014

Reputation: 462

@phoebe you can get the desired output using a slightly modified approach 

df %>% filter (!(ID %in% bad) & tag != "no good")

Generating test dataframe df

bad <- c(101, 103, 107, 110)
df  <- data.frame(ID=c(101:110),tag = rep(c("good","no good"),5))
> df
    ID     tag
1  101    good
2  102 no good
3  103    good
4  104 no good
5  105    good
6  106 no good
7  107    good
8  108 no good
9  109    good
10 110 no good

Checking output

library(dplyr)
> df %>% filter (!(ID %in% bad) & tag != "no good")
   ID  tag
1 105 good
2 109 good

Upvotes: 0

Related Questions