stopifnot {base} | R Documentation |
If any of the expressions (in ...
or exprs
) are not
all
TRUE
, stop
is called, producing
an error message indicating the first expression which was not
(all
) true.
stopifnot(..., exprs, local = TRUE)
..., exprs |
any number of (typically but not necessarily
{ expr1 expr2 .... } |
local |
(only when |
This function is intended for use in regression tests or also argument checking of functions, in particular to make them easier to read.
stopifnot(A, B)
or equivalently stopifnot(exprs= {A ;
B})
are conceptually equivalent to
{ if(any(is.na(A)) || !all(A)) stop(...); if(any(is.na(B)) || !all(B)) stop(...) }
Since R version 3.6.0, stopifnot()
no longer handles potential
errors or warnings (by tryCatch()
etc) for each single
expression
and may use sys.call(<n>)
to get a meaningful and short
error message in case an expression did not evaluate to all TRUE. This
provides considerably less overhead.
Since R version 3.5.0, expressions are evaluated sequentially,
and hence evaluation stops as soon as there is a “non-TRUE”, as
indicated by the above conceptual equivalence statement.
Further, when such an expression signals an error or
warning
, the message produced no longer
contains the full stopifnot
call, but just the erroneous
expression.
Also, since R version 3.5.0, stopifnot(exprs = { ... })
can be used
alternatively and may be preferable in the case of several
expressions, as they are more conveniently evaluated interactively
(“no extraneous ,
”).
Since R version 3.4.0, when an expression (from ...
) is not
true and is a call to all.equal
, the error
message will report the (first part of the) differences reported by
all.equal(*)
.
(NULL
if all statements in ...
are TRUE
.)
Trying to use the stopifnot(exprs = ..)
version via a shortcut,
say,
assertWRONG <- function(exprs) stopifnot(exprs = exprs)
is delicate and the above is not a good idea. Contrary to stopifnot()
which takes care to evaluate the parts of exprs
one by one and
stop at the first non-TRUE, the above short cut would typically evaluate
all parts of exprs
and pass the result, i.e., typically of the
last entry of exprs
to stopifnot()
.
However, a more careful version,
assert <- function(exprs) eval.parent(substitute(stopifnot(exprs = exprs)))
may be a nice short cut for stopifnot(exprs = *)
calls using the
more commonly known verb as function name.
stop
, warning
;
assertCondition
in package tools complements
stopifnot()
for testing warnings and errors.
stopifnot(1 == 1, all.equal(pi, 3.14159265), 1 < 2) # all TRUE m <- matrix(c(1,3,3,1), 2, 2) stopifnot(m == t(m), diag(m) == rep(1, 2)) # all(.) |=> TRUE op <- options(error = expression(NULL)) # "disabling stop(.)" << Use with CARE! >> stopifnot(all.equal(pi, 3.141593), 2 < 2, all(1:10 < 12), "a" < "b") ## More convenient for interactive "line by line" evaluation: stopifnot(exprs = { all.equal(pi, 3.1415927) 2 < 2 all(1:10 < 12) "a" < "b" }) # long all.equal() error messages are abbreviated: stopifnot(all.equal(rep(list(pi),4), list(3.1, 3.14, 3.141, 3.1415))) options(op) # revert to previous error handler