README.md

In kpien/msafer: Check If You Can Use Map

msafer

msafer is a package that frees users from having to manually find the location of an error when using the function map(). When map() returns an error, it’s unclear where this error occurs. The function map_safe() takes in a vector, function and requirements, and spits out TRUE or FALSE based on whether an error occurs when running the function over the vector. map_safe() will return a list of logical returns where it will be easy to identify the place of the error.

Installation

You can install the development version of msafer from GitHub with:

# install.packages("devtools")
devtools::install_github("kpien/msafer")

Example

For demonstration purposes, let’s create a sample list of dataframes from the starwars and mtcars datasets.

library(dplyr)
sample_a <- starwars %>%
  sample_n(34) %>%
  select(-height, -hair_color)
sample_b <- sample_n(starwars, 35)
sample_c <- mtcars %>%
  sample_n(20) %>%
  select(-hp)
sample_d <- tibble::tibble()
list_sample <- list(starwars, sample_a, sample_b, sample_c, mtcars, sample_d)

Pass map_safe_merge() the same arguments as map(). These would be the list of dataframes, the function to map over all the dataframes, and any arguments that the function needs. map_safe_merge() will return a tibble with the file numbers and any errors that may have occurred while trying to apply map(), grouping repeat errors together.

library(msafer)
map_safe_merge(list_sample, select, height)
#> # A tibble: 6 x 2
#>   result error_message                                           
#>   <lgl>  <chr>                                                   
#> 1 TRUE   NA                                                      
#> 2 FALSE  "Error in .f(.x[[i]], ...): object 'height' not found\n"
#> 3 TRUE   NA                                                      
#> 4 FALSE  "Error in .f(.x[[i]], ...): object 'height' not found\n"
#> 5 FALSE  "Error in .f(.x[[i]], ...): object 'height' not found\n"
#> 6 FALSE  "Error in .f(.x[[i]], ...): object 'height' not found\n"

However, if the vector is huge, then the result would be messy and hard to grab information from the tibble. Thus, we can use map_safe() which nests the error message, and returns a tibble that contains only the unique error message and the index locating where the error occurs within the vector.

sample_df <- map_safe(list_sample, dplyr::select, height)
sample_df
#> # A tibble: 2 x 3
#>   result error_message                                                  which_id
#>   <lgl>  <chr>                                                    <list<df[,1]>>
#> 1 TRUE   NA                                                              [2 × 1]
#> 2 FALSE  "Error in .f(.x[[i]], ...): object 'height' not found\n"        [4 × 1]
sample_df %>% 
  tidyr::unnest(which_id)
#> # A tibble: 6 x 3
#>   result error_message                                               id
#>   <lgl>  <chr>                                                    <int>
#> 1 TRUE   NA                                                           1
#> 2 TRUE   NA                                                           3
#> 3 FALSE  "Error in .f(.x[[i]], ...): object 'height' not found\n"     2
#> 4 FALSE  "Error in .f(.x[[i]], ...): object 'height' not found\n"     4
#> 5 FALSE  "Error in .f(.x[[i]], ...): object 'height' not found\n"     5
#> 6 FALSE  "Error in .f(.x[[i]], ...): object 'height' not found\n"     6

check_match() identified whether the user’s requirement existed within the dataset/list of dataframes, if yes, then the function will return TRUE, if not, it returns FALSE.

check_match(dplyr::starwars, height == 172)
#> [1] TRUE

check_match() and also be used under map_safe() when applying check_match() to a vector which could contain multiple elements

d <- map_safe(list_sample, check_match,  height == 172)
d
#> # A tibble: 2 x 3
#>   result error_message                              which_id
#>   <lgl>  <chr>                                <list<df[,1]>>
#> 1 TRUE   NA                                          [3 × 1]
#> 2 FALSE  "Error: object 'height' not found\n"        [3 × 1]

Developed By

• Sifan(Carol) Liu
• Kelly Pien
• Huiqing(Lily) Jin

kpien/msafer documentation built on Dec. 25, 2019, 5:12 a.m.