checkinput

With checkinput, you can write concise, flexible checks for input to R functions.

Installation

Visit the checkinput website to explore the package, or install checkinput from GitHub using the following R code:

if(!requireNamespace("remotes")) {
  install.packages(pkgs = "remotes")
}
remotes::install_github(repo = "JesseAlderliesten/checkinput",
                        upgrade = FALSE, build_vignettes = TRUE, lib = NULL)

For more information about installing and configuring R, see my package checkrpkgs.

Example

Say you want to collect information on people’s hobbies through an R function in which it is optional for them to provide their name. With checkinput, you could write a function like list_hobbies():

list_hobbies <- function(name, age, hobbies) {
  stopifnot(is_character(name, allow_empty = TRUE, allow_NA = TRUE),
            is_nonnegative(age), all_characters(hobbies))
  list(name = name, age = age, hobbies = hobbies)
}

The checks inside stopifnot() ensure that name contains a single character string that might be empty ("") or NA; age contains a single non-negative number; hobbies contains at least one character string and does not contain empty strings or NAs.

The base R equivalent of list_hobbies() would require much more code to check the input, increasing the chance of coding errors and making it more difficult to read:

list_hobbies_base <- function(name, age, hobbies) {
  stopifnot(is.character(name), is.atomic(name), is.null(dim(name)),
            length(name) == 1L,
            is.numeric(age), is.atomic(age), is.null(dim(age)),
            length(age) == 1L, !is.na(age), age >= 0L,
            is.character(hobbies), is.atomic(hobbies), is.null(dim(hobbies)),
            length(hobbies) > 0L, all(nzchar(hobbies, keepNA = FALSE)),
            !anyNA(hobbies))
  list(name = name, age = age, hobbies = hobbies)
}

If the input passes all checks, both functions produce the same output:

library(checkinput)
hobbies_John <- c("books", "construction sets")
John_checkinput <- list_hobbies(name = "John", age = 25, hobbies = hobbies_John)
John_base <- list_hobbies_base(name = "John", age = 25, hobbies = hobbies_John)
identical(John_checkinput, John_base)
#> [1] TRUE

hobbies_baby <- "drinking milk"
baby_checkinput <- list_hobbies(name = "", age = 0, hobbies = hobbies_baby)
baby_base <- list_hobbies_base(name = "", age = 0, hobbies = hobbies_baby)
identical(baby_checkinput, baby_base)
#> [1] TRUE

When a check fails, error messages indicate the offending arguments:

library(checkinput)
try(list_hobbies(name = "John", age = 25, hobbies = c(hobbies_John, "")))
#> Error in list_hobbies(name = "John", age = 25, hobbies = c(hobbies_John,  : 
#>   all_characters(hobbies) is not TRUE
try(list_hobbies(name = "", age = -1, hobbies = hobbies_baby))
#> Error in list_hobbies(name = "", age = -1, hobbies = hobbies_baby) : 
#>   is_nonnegative(age) is not TRUE

Design choices

The check functions of checkinput return either TRUE or FALSE and do not throw errors for any input to x. Errors are thrown about invalid input to arguments other than x, e.g., when values other than TRUE or FALSE are used for argument allow_NA. This is explained in the vignette about design choices: vignette("design_choices", package = "checkinput"). That vignette also shows how to get a named boolean vector indicating for each element of x if it is TRUE or FALSE according to the check functions.

checkinput also contains a vignette about type coercion in relation to checking function input:
vignette("type_coercion", package = "checkinput").

Function overview

The check functions in checkinput can be divided into groups based on the kind of input for which they return TRUE:

• zero-length values: is_zerolength()
• logical values: is_logical()
• characters: is_character(), all_characters(), is_path(), all_names()
• numbers: is_number(), all_numbers(), is_nonnegative(), all_nonnegative(), is_positive(), is_natural(), all_natural()

Besides these check functions, checkinput contains utility-functions:

• make_natural() which returns a natural number if the input is nearly equal to such a number
• paste_quoted() to concatenate input to a character string with quoted elements

License

This project is licensed under the terms of the MIT License.

Citation

To cite package 'checkinput' in publications use:

  Alderliesten J (2026). _checkinput: Check Function Input_. R package
  version 1.5.1, <https://github.com/JesseAlderliesten/checkinput>.

A BibTeX entry for LaTeX users is

  @Manual{,
    title = {checkinput: Check Function Input},
    author = {Jesse Alderliesten},
    year = {2026},
    note = {R package version 1.5.1},
    url = {https://github.com/JesseAlderliesten/checkinput},
  }

Contact

Please open a GitHub issue if you have suggestions to improve checkinput.

Similar packages

Functions of checkinput use arguments to determine if special values like NA and zero-length values should be allowed, making them more flexible than functions in similar R packages that require different functions to include these special values. Nevertheless, the following similar packages are worth looking into:

• arkhe: tools for cleaning rectangular data
• assertable: verbose assertions for tabular data (data.frames and data.tables)
• assertthat: easy pre and post assertions
• checkmate: fast and versatile argument checks
• chk: check user-supplied function arguments
• erify: check arguments and generate readable error messages