[R] stopifnot with logical(0)

Martin Maechler maechler at stat.math.ethz.ch
Mon Dec 14 11:10:41 CET 2015

>>>>> "DM" == Duncan Murdoch <murdoch.duncan at gmail.com>
>>>>>     on Sat, 12 Dec 2015 09:05:04 -0500 writes:

    DM> On 12/12/2015 8:44 AM, peter dalgaard wrote:
    >>> On 12 Dec 2015, at 10:54 , Martin Maechler
    >>> <maechler at stat.math.ethz.ch> wrote:
    >>> My conclusion: Breaking such a fundamental lemma of
    >>> logic as "the empty set is always true"
    >> Umm, that doesn't make sense to me. Surely you mean that
    >> "an AND-operation over an empty index set is TRUE"? A
    >> similar OR operation is FALSE, i.e. they behave like
    >> empty products and sums, respectively.

    DM> How about "the empty set is all true, and all false."

or, what the I *meant* with the above:

  "All statements about elements of the empty set are true"

  ((and I still like the short form, even though it is not
    correct strictly logically/mathematically))
Of course, Peter is correct,  and that
   any(logical(0))   is  FALSE
is really the only sensical way.

More information about the R-help mailing list