Delete a row in R based on next row condition

Question

I'm having the following problem:

I have a large dataframe for which I'd like to delete every row that matches this condition: if a value (string) inside a column contains the ":" char and the next row also contains a ":" char, it would delete the first one.

Something like this:

a <- c("value1","value2","value2:a","value2:b","value3")
b <- c(1,2,3,4,5)
df1 <- data.frame(b,a)

  b        a
1 1   value1
2 2   value2
3 3 value2:a
4 4 value2:b
5 5   value3

to this, where only the row containing the name "value2:a" gets deleted because is followed by a row that contains the regex ":"

a        b
value1   1
value2   2
value2:b 4
value3   5

Thank you very much in advance, I've been trying out some solutions with for loops and the grepl function but can't seem to make it work.

Answer 1

You don't need loops for this. You can do it with a single line in base R by combining a grepl with a lagged grepl

df1[!c(head(grepl(":", df1$a), -1) & tail(grepl(":", df1$a), -1), FALSE),]
#>   b        a
#> 1 1   value1
#> 2 2   value2
#> 4 4 value2:b
#> 5 5   value3

Answer 2

Try this:

df1 %>% 
    dplyr::filter(
        !(stringr::str_detect(a, ":") & stringr::str_detect(lead(a), ":"))
    )

Answer 3

Here is a dplyr solution.

library(dplyr)

df1 %>%
  mutate(flag = grepl(":", a),
         flag = cumsum(flag)*flag,
         flag = lead(flag, default = 0)) %>%
  filter(flag != 2) %>%
  dplyr::select(-flag)
#  b        a
#1 1   value1
#2 2   value2
#3 4 value2:b
#4 5   value3

Answer 4

Here you go. For tasks like this I always like to use a couple of masks.

a <- c("value1","value2","value2:a","value2:b","value3")
b <- c(1,2,3,4,5)
df1 <- as.data.frame(b,a)
stringr::str_extract(pattern = ":",string = rownames(df1)) -> vec
mask_colon = duplicated(vec,fromLast = FALSE)
mask_na = is.na(vec)
df1 = df1[which(mask_na | mask_colon),, drop = FALSE]
df1
#>          b
#> value1   1
#> value2   2
#> value2:b 4
#> value3   5

^{Created on 2020-10-07 by the reprex package (v0.3.0)}