Remove rows in df using multiple conditions in R

Question

Is it possible to remove rows of data by referencing specific character strings or factor levels from 2 or more columns? For small datasets, this is easy because I can just scroll through the dataframe and remove the row I want, but how could this be achieved for larger datasets without endlessly scrolling to see which rows match my criteria?

Fake data:

df1 <- data.frame(year = rep(c(2019, 2020), each = 10),
                  month = rep(c("March", "October"), each = 1), 
                  site = rep(c("1", "2", "3", "4", "5"), each = 2),
                  common_name = rep(c("Tuna", "shark"), each = 1),
                  num = sample(x = 0:2, size  = 20, replace = TRUE))

For example: How do I remove only site "1" in March of 2019 in one line of code and without looking at which row it's in?

Also see [here](https://stackoverflow.com/questions/8005154/subset-data-frame-based-on-multiple-conditions) and [here](https://stackoverflow.com/questions/6601658/deleting-specific-rows-from-a-data-frame). — DaveArmstrong, May 31 '22 at 20:13
Does this answer your question? [Deleting specific rows from a data frame](https://stackoverflow.com/questions/6601658/deleting-specific-rows-from-a-data-frame) — DaveArmstrong, May 31 '22 at 20:14

score 2 · Accepted Answer · answered May 31 '22 at 20:06

You can use subset():

df1 <- data.frame(year = rep(c(2019, 2020), each = 10),
                  month = rep(c("March", "October"), each = 1), 
                  site = rep(c("1", "2", "3", "4", "5"), each = 2),
                  common_name = rep(c("Tuna", "shark"), each = 1),
                  num = sample(x = 0:2, size  = 20, replace = TRUE))

subset(df1, !(site == "1" & year == 2019 & month == "March"))
#>    year   month site common_name num
#> 2  2019 October    1       shark   0
#> 3  2019   March    2        Tuna   1
#> 4  2019 October    2       shark   0
#> 5  2019   March    3        Tuna   0
#> 6  2019 October    3       shark   0
#> 7  2019   March    4        Tuna   2
#> 8  2019 October    4       shark   2
#> 9  2019   March    5        Tuna   0
#> 10 2019 October    5       shark   2
#> 11 2020   March    1        Tuna   1
#> 12 2020 October    1       shark   1
#> 13 2020   March    2        Tuna   2
#> 14 2020 October    2       shark   2
#> 15 2020   March    3        Tuna   1
#> 16 2020 October    3       shark   0
#> 17 2020   March    4        Tuna   1
#> 18 2020 October    4       shark   0
#> 19 2020   March    5        Tuna   0
#> 20 2020 October    5       shark   2

^{Created on 2022-05-31 by the reprex package (v2.0.1)}

score 1 · Answer 2 · answered May 31 '22 at 20:13

We could use paste as well

subset(df1, paste(year, month, site) != '2019 March 1')

-output

   year   month site common_name num
2  2019 October    1       shark   1
3  2019   March    2        Tuna   1
4  2019 October    2       shark   2
5  2019   March    3        Tuna   0
6  2019 October    3       shark   0
7  2019   March    4        Tuna   2
8  2019 October    4       shark   1
9  2019   March    5        Tuna   1
10 2019 October    5       shark   1
11 2020   March    1        Tuna   1
12 2020 October    1       shark   1
13 2020   March    2        Tuna   1
14 2020 October    2       shark   2
15 2020   March    3        Tuna   1
16 2020 October    3       shark   0
17 2020   March    4        Tuna   1
18 2020 October    4       shark   1
19 2020   March    5        Tuna   1
20 2020 October    5       shark   2

Why does filter not work here? `dplyr::filter(df1, (site !="1" | month !="March" | year !=2019))` — TarJae, May 31 '22 at 20:14
@TarJae It should work `dplyr::filter(df1, paste(year, month, site) != '2019 March 1')` — akrun, May 31 '22 at 20:15

score 1 · Answer 3 · answered May 31 '22 at 20:23

1

A one line alternative to subset or dplyr:filter using the R bracket notation:

df2 <- df1[!(df1$site=="1" & df1$year==2019 & df1$month=="March"),]

answered May 31 '22 at 20:23

M.Viking

5,067
4
17
33

Remove rows in df using multiple conditions in R

3 Answers3