delete row if there is a specific word

[dario_gd]

If I have this dataframe

prove<- data.frame(a=c(1,2,3),
                   b=c("test_4", "test_4fail", "test_5"),
                   c=c(8,4,5))

What is the most efficient way to delete rows where in column b, there is the word "Fail"

[mfherman]

You could use a combination of dplyr::filter() and stringr::str_detect(). Note, I use the negate = TRUE argument of str_detect() to indicate I want to keep rows where "fail" does not appear.

library(tidyverse)

prove <- data.frame(
  a = c(1, 2, 3),
  b = c("test_4", "test_4fail", "test_5"),
  c = c(8, 4, 5)
)

prove %>%
  filter(str_detect(b, "fail", negate = TRUE))
#>   a      b c
#> 1 1 test_4 8
#> 2 3 test_5 5

^{Created on 2019-02-20 by the reprex package (v0.2.1)}

[dario_gd]

i have this ERROR

Error in filter_impl(.data, quo) : 
  Evaluation error: unused argument (negate = T).

[mfherman]

Looks like you have an old version of stringr installed. The negate argument was added in version 1.4.0. You could update to the latest version of stringr by running:

install.packages("stringr")

Alternatively, you could use ! to negate str_detect() like so:

library(tidyverse)

prove <- data.frame(
  a = c(1, 2, 3),
  b = c("test_4", "test_4fail", "test_5"),
  c = c(8, 4, 5)
)

prove %>%
  filter(!str_detect(b, "fail"))
#>   a      b c
#> 1 1 test_4 8
#> 2 3 test_5 5

^{Created on 2019-02-20 by the reprex package (v0.2.1)}

[andresrcs]

This question and your previous questions are highly correlated with patterns in strings, so this information could be very useful for you to learn more about this topic.

https://stringr.tidyverse.org/articles/regular-expressions.html