fbpx
R Tip: Use isTRUE() R Tip: Use isTRUE()
R Tip: use isTRUE(). A lot of R functions are type unstable, which means they return different types or classes depending on details of their values. For example... R Tip: Use isTRUE()

R Tip: use isTRUE().

A lot of R functions are type unstable, which means they return different types or classes depending on details of their values.

For example consider all.equal(), it returns the logical value TRUEwhen the items being compared are equal:

all.equal(1:3, c(1, 2, 3))
# [1] TRUE

However, when the items being compared are not equal all.equal() instead returns a message:

all.equal(1:3, c(1, 2.5, 3))
# [1] "Mean relative difference: 0.25"

This can be inconvenient in using functions similar to all.equal()as tests in if()-statements and other program control structures.

The saving functions is isTRUE()isTRUE() returns TRUE if its argument value is equivalent to TRUE, and returns FALSE otherwise.isTRUE() makes R programming much easier.

Some examples of isTRUE() are given below:

isTRUE(TRUE)
# [1] TRUE

isTRUE(FALSE)
[1] FALSE

isTRUE(NULL)
# [1] FALSE

isTRUE(NA)
# [1] FALSE

isTRUE(all.equal(1:3, c(1, 2.5, 3)))
# [1] FALSE

isTRUE(all.equal(1:3, c(1, 2, 3)))
# [1] TRUE

lst <- list(x = 5)
isTRUE(lst$y == 7)
# [1] FALSE
lst$y == 7

logical(0)
isTRUE(logical(0))
# [1] FALSE

Using isTRUE() one can write safe and legible code such as the following:

# Pretend this assignment was performed by somebody else.
lst <- list(x = 5)

# Our own sanitization code.
if(!isTRUE(lst$y > 3)) {
  lst$y = lst$x
}

print(lst)

# $x
# [1] 5
#
# $y
# [1] 5

R now also has isFALSE(), but by design it does not mean the same thing as !isTRUE(). The ideas is: for a value v at most of one of isTRUE() or isFALSE() is set, and both are non-NA unnamed scalar logical values. (example: isTRUE(5)isFALSE(5)).

Or as help(isTRUE) puts it:

… if(isTRUE(cond)) may be preferable to if(cond) …

Note: prior to R 3.5 isTRUE (the current version!) was defined as “isTRUE <- function(x) identical(x, TRUE)” (please see change-log here). This seemed clever, but failed on named logical values (violating a principle of least surprise):

oldTRUE <- function(x) identical(x, TRUE)

v <- TRUE
oldTRUE(v)
# [1] TRUE
isTRUE(v)
# [1] TRUE

names(v) <- "condition"
oldTRUE(v)
# [1] FALSE
isTRUE(v)
# [1] TRUE

This caused a lot of problems (example taken from R3.5.0 NEWS):

x <- rlnorm(99)

isTRUE(median(x) == quantile(x)["50%"])
# [1] TRUE

oldTRUE(median(x) == quantile(x)["50%"])
# [1] FALSE

Original Source

John Mount

John Mount

My specialty is analysis and design of algorithms, with an emphasis on efficient implementation. I work to find applications of state of the art methods in optimization, statistics and machine learning in various application areas. Currently co-authoring "Practical Data Science with R"

1