Replace loop with more functional code

[RayStick]

Suggestion that you do not have to tackle now since reviewers might have an opinion too.

Could the for loop be replaced with more functional code where the name of functions might help understand what's being calculated.

In R/data_manipulation.R:

is_empty_description <- function(index, table) {
  nchar(table$description[index] ||
  table$description[index] == "Description to follow") ||
  table$description[index] == "NA" # why doesn't  `is.na()` work by the way?
}
# untested, and might be wrong
table$description <- purrr::map_lgl(seq_len(nrow(table_df)), is_empty_description, table = table)
# this returns a logical not character

Originally posted by @maelle in #167 (comment)

[maelle]

Just thought that

table$description[index] == "Description to follow") ||
  table$description[index] == "NA"

could be

table$description[index] %in% c("Description to follow", "NA")

which is vectorized

descriptions <- c(
  "real one", 
  "1", 
  "a",
  "-",
  "NA", 
  "Description to follow"
)
descriptions %in% c("NA", "Description to follow")
#> [1] FALSE FALSE FALSE FALSE  TRUE  TRUE

^{Created on 2024-12-18 with reprex v2.1.1}

To also have the condition on the number of characters:

descriptions <- c(
  "real one", 
  "1", 
  "a",
  "-",
  "NA", 
  "Description to follow"
)

(nchar(descriptions) == 1) | descriptions %in% c("NA", "Description to follow")
#> [1] FALSE  TRUE  TRUE  TRUE  TRUE  TRUE

^{Created on 2024-12-18 with reprex v2.1.1}

[maelle]

So I think something like this should work:

table$empty <- (nchar(table_df$description) == 1) | table_df$description %in% c("NA", "Description to follow")

[RayStick]

See this line here:

mapmetadata/R/empty_count.R

Line 15 in b27250f

empty_count = sum((nchar(as.character(Column.description)) <= 1) | Column.description %in% c("NA", "Description to follow", NA)),

The code structure was changed a bit but I still used your suggestion as inspiration to simplify 😸

[maelle]

Yay, dplyr for the win! 🎉