mau...@alice.it wrote:
> I cannot find any R function or operator that performs a binary AND
> operation, as performed by Fortran built-in function "iand".
> Ideally either R operator "&" or "&&" should do that. But some tests proved
> they do not:
>
they do not, it seems clear from the documentation. here's a hint:
RSiteSearch('bitwise')
your question points me to another issue (thanks!):
one = as.raw(1)
as.logical(one)
# TRUE
two = as.raw(2)
as.logical(two)
# TRUE
one & two
# 00
if (one) 1 else 0
# error: unimplemented type 'raw' in 'asLogical'
oops...
how come 01 AND 02 = 00?? well, you see, & is a *logical* (not a
*bitwise*) AND operator (see ?'&'), but for raws it does bitwise AND
(see ?'&' again):
as.raw(6) & as.raw(3)
# 02
?'&' says:
"
Arguments:
x, y: logical vectors, or objects which can be coerced to such or
for which methods have been written.
"
this might have been more explicit about raws, but the use of logical
operators (that is, bitwise operators) with raws *is* described in the
details.
the other case is weird; ?'if' says:
"
Arguments:
cond: A length-one logical vector that is not 'NA'. Conditions of
length greater than one are accepted with a warning, but only
the first element is used. Other types are coerced to
logical if possible, ignoring any class.
"
and the help page does not mention 'raw' at all. raws are not logical:
is.logical(one)
# FALSE
(what about the type hierarchy???), yet it is possible to convert them
to logical:
as.logical(one)
# TRUE
why if(one) chooses to raise an error is a mystery to me. the null
hypothesis: a design flaw. the alternative one: a bug. choose your test.
best,
vQ
______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
