how to filter for a string across columns one at a time-CodePudding

In the table below, I am trying to filter out rows with the word blue. The final result should be two rows for 11th and 12th oct. What I have tried which is not working:

tibble::tribble(
         ~date, ~col_1, ~col_2, ~col_3,
  "11/10/2021", "blue",  "jhf",  "fff",
  "12/10/2021",  "ert", "blue",  "kkk",
  "13/10/2021",  "yui",  "yui", "blkeee"
  ) %>% 
  
  mutate(date = lubridate::dmy(date)) %>% 
  
  dplyr::filter(across(.cols = where(is.character), 
                .fns = ~str_detect(., regex("blue", ignore_case = TRUE))
  ))

Any ideas?

CodePudding user response：

You may use if_any -

library(dplyr)
library(stringr)

df %>%
  filter(if_any(starts_with('col'), 
                ~str_detect(., regex("blue", ignore_case = TRUE))))

#  date       col_1 col_2 col_3
#  <chr>      <chr> <chr> <chr>
#1 11/10/2021 blue  jhf   fff  
#2 12/10/2021 ert   blue  kkk

In base R -

subset(df, Reduce(`|`, lapply(df[-1], grepl, pattern = "blue", ignore.case = TRUE)))

CodePudding user response：

# Base R option 1:
df[rowSums(Vectorize(`==`)("blue", df)) >= 1,]

# Base R option 2:
df[
  apply(
    df[,
       vapply(
         df, 
         is.character,
         logical(1)
         )
      ], 
    1, 
    function(x){
      any(x == "blue")
    }
  ),
]