cast_safe_df()

Safely run a function and return a consistent data frame Executes any function while handling errors and empty outputs gracefully. Returns a tibble with the original results, plus context and error columns.

Usage

cast_safe_df(
  fun,
  ...,
  context = NULL,
  placeholder = tibble::tibble(result = NA)
)

Arguments

fun

A function to execute.

...

Arguments passed to fun.

context

Optional character string describing the context of this call (e.g., filename or input value). Defaults to NULL.

placeholder

A tibble to return if the function fails or returns an empty data frame. Defaults to tibble(result = NA).

Value

A tibble containing:

result

The output of the function (or placeholder if failed).

context

The context string provided.

error

The error message if the function failed, otherwise NA.

Details

This function is useful for safely executing functions that may fail on some inputs, such as reading multiple files where some may be missing or malformed. It ensures that the output is always a tibble and can be safely combined using purrr::map_df().

Examples

library(tibble)
library(purrr)
library(readr)

# Example 1: math function with possible errors
f <- function(x) {
  if (x == 0) stop("Cannot divide by zero!")
  10 / x
}
nums <- -1:1
out <- map_df(nums, ~ cast_safe_df(
  f, .x,
  context = paste("value =", .x),
  placeholder = tibble(result = NA_real_)
))
#> Warning: Error in value = 0: Cannot divide by zero!
print(out)
#> # A tibble: 3 × 3
#>   result context    error                 
#>    <dbl> <chr>      <chr>                 
#> 1    -10 value = -1 NA                    
#> 2     NA value = 0  Cannot divide by zero!
#> 3     10 value = 1  NA                    

# Example 2: reading multiple files safely
df <- map_df(list.files(pattern = ".csv", full.names = TRUE),
             ~ cast_safe_df(readr::read_csv, .x,
                           col_names = FALSE,
                           show_col_types = FALSE,
                           context = basename(.x),
                           placeholder = tibble(X1 = NA_character_)))