When you write functions that operate on S3 or unclassed objects you can either trust that your inputs will be structured as expected, or tediously check that they are.
vetr takes the tedium out of structure verification so that you can
trust, but verify. It lets you express structural requirements
declaratively with templates, and it auto-generates human-friendly error
messages as needed.
vetr is written in C to minimize overhead from parameter checks in
your functions. It has no dependencies.
Declare a template that an object should conform to, and let vetr take
care of the rest:
library(vetr)
tpl <- numeric(1L)
vet(tpl, 1:3)
## [1] "`length(1:3)` should be 1 (is 3)"
vet(tpl, "hello")
## [1] "`\"hello\"` should be type \"numeric\" (is \"character\")"
vet(tpl, 42)
## [1] TRUE
The template concept is based on vapply, but generalizes to all S3
objects and adds some special features to facilitate comparison. For
example, zero length templates match any length:
tpl <- integer()
vet(tpl, 1L:3L)
## [1] TRUE
vet(tpl, 1L)
## [1] TRUE
And for convenience short (<= 100 length) integer-like numerics are considered integer:
tpl <- integer(1L)
vet(tpl, 1) # this is a numeric, not an integer
## [1] TRUE
vet(tpl, 1.0001)
## [1] "`1.0001` should be type \"integer-like\" (is \"double\")"
vetr can compare recursive objects such as lists, or data.frames:
tpl.iris <- iris[0, ] # 0 row DF matches any number of rows in object
iris.fake <- iris
levels(iris.fake$Species)[3] <- "sibirica" # tweak levels
vet(tpl.iris, iris)
## [1] TRUE
vet(tpl.iris, iris.fake)
## [1] "`levels(iris.fake$Species)[3]` should be \"virginica\" (is \"sibirica\")"
From our declared template iris[0, ], vetr infers all the required
checks. In this case, vet(iris[0, ], iris.fake, stop=TRUE) is
equivalent to:
stopifnot_iris <- function(x) {
stopifnot(
is.data.frame(x),
is.list(x),
length(x) == length(iris),
identical(lapply(x, class), lapply(iris, class)),
is.integer(attr(x, 'row.names')),
identical(names(x), names(iris)),
identical(typeof(x$Species), "integer"),
identical(levels(x$Species), levels(iris$Species))
)
}
stopifnot_iris(iris.fake)
## Error in stopifnot_iris(iris.fake): identical(levels(x$Species), levels(iris$Species)) is not TRUE
vetr saved us typing, and the time and thought needed to come up with
what needs to be compared.
You could just as easily have created templates for nested lists, or
data frames in lists. Templates are compared to objects with the alike
function. For a thorough description of templates and how they work see
the alike
vignette.
For template examples see example(alike).
Let’s revisit the error message:
vet(tpl.iris, iris.fake)
## [1] "`levels(iris.fake$Species)[3]` should be \"virginica\" (is \"sibirica\")"
It tells us:
- The reason for the failure
- What structure would be acceptable instead
- The location of failure
levels(iris.fake$Species)[3]
vetr does what it can to reduce the time from error to resolution. The
location of failure is generated such that you can easily copy it in
part or full to the R prompt for further examination.
You can combine templates with && / ||:
vet(numeric(1L) || NULL, NULL)
## [1] TRUE
vet(numeric(1L) || NULL, 42)
## [1] TRUE
vet(numeric(1L) || NULL, "foo")
## [1] "`\"foo\"` should be `NULL`, or type \"numeric\" (is \"character\")"
Templates only check structure. When you need to check values use . to
refer to the object:
vet(numeric(1L) && . > 0, -42) # strictly positive scalar numeric
## [1] "`-42 > 0` is not TRUE (FALSE)"
vet(numeric(1L) && . > 0, 42)
## [1] TRUE
If you do use the . symbol in your vetting expressions in your
packages, you will need to include utils::globalVariables(".") as a
top-level call to avoid the “no visible binding for global variable ‘.’”
R CMD check NOTE.
You can compose vetting expressions as language objects and combine them:
scalar.num.pos <- quote(numeric(1L) && . > 0)
foo.or.bar <- quote(character(1L) && . %in% c('foo', 'bar'))
vet.exp <- quote(scalar.num.pos || foo.or.bar)
vet(vet.exp, 42)
## [1] TRUE
vet(vet.exp, "foo")
## [1] TRUE
vet(vet.exp, "baz")
## [1] "At least one of these should pass:"
## [2] " - `\"baz\" %in% c(\"foo\", \"bar\")` is not TRUE (FALSE)"
## [3] " - `\"baz\"` should be type \"numeric\" (is \"character\")"
all_bw is available for value range checks (~10x faster than
isTRUE(all(. >= x & . <= y)) for large vectors):
vet(all_bw(., 0, 1), runif(5) + 1)
## [1] "`all_bw(runif(5) + 1, 0, 1)` is not TRUE (is chr: \"`1.614709` at index 1 not in `[0,1]`\")"
There are a number of predefined vetting tokens you can use in your vetting expressions such as:
vet(NUM.POS, -runif(5)) # positive numeric; see `?vet_token` for others
## [1] "`-runif(5)` should contain only positive values, but has negatives"
Vetting expressions are designed to be intuitive to use, but their
implementation is complex. We recommend you look at example(vet) for
usage ideas, or at the “Non Standard Evaluation” section of the
vignette
for the gory details.
If you are vetting function inputs, you can use the vetr function,
which works just like vet except that it is streamlined for use within
functions:
fun <- function(x, y) {
vetr(numeric(1L), logical(1L))
TRUE # do work...
}
fun(1:2, "foo")
## Error in fun(x = 1:2, y = "foo"): For argument `x`, `length(1:2)` should be 1 (is 2)
fun(1, "foo")
## Error in fun(x = 1, y = "foo"): For argument `y`, `"foo"` should be type "logical" (is "character")
vetr automatically matches the vetting expressions to the
corresponding arguments and fetches the argument values from the
function environment.
See
vignette
for additional details on how the vetr function works.
vetrvignette,?vet,?vetr,example(vet),example(vetr).alikevignette,?alike, andexample(alike)for discussion of templates.- A survey of parameter validation functions.
vetr is still in development, although most of the features are
considered mature. The most likely area of change is the treatment of
function and language templates (e.g. alike(sum, max)), and more
flexible treatment of list templates (e.g. in future lists may be
allowed to be different lengths so long as every named element in the
template exists in the object).
This package is available on CRAN:
install.packages('vetr')
It has no runtime dependencies.
For the development version use
remotes::install_github('brodieg/vetr@development') or:
f.dl <- tempfile()
f.uz <- tempfile()
github.url <- 'https://github.com/brodieG/vetr/archive/development.zip'
download.file(github.url, f.dl)
unzip(f.dl, exdir=f.uz)
install.packages(file.path(f.uz, 'vetr-development'), repos=NULL, type='source')
unlink(c(f.dl, f.uz))
The master branch typically mirrors CRAN and should be stable.
There are many alternatives available to vetr. We do a survey of the
following in our parameter validation
functions
review:
stopifnotby R Corevetrby Yours Trulyasserthatby Hadley Wickhamassertiveby Richie Cottoncheckmateby Michel Lang
The following packages also perform related tasks, although we do not review them:
valaddinv0.1.0 by Eugene Ha, a framework for augmenting existing functions with validation contracts. Currently the package is undergoing a major overhaul so we will add it to the comparison once the new release (v0.3.0) is out.ensurerv1.1 by Stefan M. Bache, a framework for flexibly creating and combining validation contracts. The development version adds an experimental method for creating type safe functions, but it is not published to CRAN so we do not test it here.validateby Mark van der Loo and Edwin de Jonge, with a primary focus on validating data in data frames and similar data structures.assertrby Tony Fischetti, also focused on data validation in data frames and similar structures.typesby Jim Hester, which implements but does not enforce type hinting.argufyby Gábor Csárdi, which implements parameter validation via roxygen tags (not released to CRAN).typedby Antoine Fabri, which enforces types of symbols, function parameters, and return values.
Thank you to:
- R Core for developing and maintaining such a wonderful language.
- CRAN maintainers, for patiently shepherding packages onto CRAN and maintaining the repository, and Uwe Ligges in particular for maintaining Winbuilder.
- Users and others who have reported bugs and/or helped contribute fixes (see NEWS.md).
- Tomas Kalibera for rchk and rcnst to help detect errors in compiled code, and in particular for his infinite patience in helping me resolve the issues he identified for me.
- Jim Hester because covr rocks.
- Dirk Eddelbuettel and Carl Boettiger for the rocker project, and Gábor Csárdi and the R-consortium for Rhub, without which testing bugs on R-devel and other platforms would be a nightmare.
- Winston Chang for the r-debug docker container, in particular because of the valgrind level 2 instrumented version of R.
- Hadley Wickham and Peter Danenberg for roxygen2.
- Yihui Xie for knitr and J.J. Allaire etal for rmarkdown, and by extension John MacFarlane for pandoc.
- Michel Lang for pushing me to implement
all_bwto compete with his own packagecheckmate. - Eugene Ha for pointing me to several other relevant packages, which in turn led to the survey of related packages.
- Stefan M. Bache for the idea of having a function for testing
objects directly (originally
vetronly worked with function arguments), which I took from ensurer. - Olaf Mersmann for microbenchmark, because microsecond matter, and Joshua Ulrich for making it lightweight.
- All open source developers out there that make their work freely available for others to use.
- Github, Codecov, Vagrant, Docker, Ubuntu, Brew for providing infrastructure that greatly simplifies open source development.
- Free Software Foundation for developing the GPL license and promotion of the free software movement.
Brodie Gaslam is a hobbyist programmer based on the US East Coast.