today <- Sys.Date() typeof(today) # [1] "double" typeof(seq(today, by=1, length.out=2)) # [1] "integer" Clearly minor as it doesn't seem to have come up before (e.g. coercion to numeric will happen automatically whenever fractional dates are needed); I only noticed because of a test using identical failing: identical( seq(today, by=1, length.out=10), today + 0:9 ) # [1] FALSE It's easy in this case to fix the test using coercion, but this could (and did in practice) come up at deeper levels of nesting that become more onerous to handle. And using all.equal() comes with its own tradeoffs. The fix would be easy enough -- at a glance there are two usages of .Date(seq.int(...)) in seq.Date() that could be replaced by .Date(as.numeric(seq.int(...))). Mike C
>>>>> Michael Chirico via R-devel writes:> today <- Sys.Date() > typeof(today) > # [1] "double" > typeof(seq(today, by=1, length.out=2)) > # [1] "integer"> Clearly minor as it doesn't seem to have come up before (e.g. coercion > to numeric will happen automatically whenever fractional dates are > needed); I only noticed because of a test using identical failing:> identical( > seq(today, by=1, length.out=10), > today + 0:9 > ) > # [1] FALSE> It's easy in this case to fix the test using coercion, but this could > (and did in practice) come up at deeper levels of nesting that become > more onerous to handle. And using all.equal() comes with its own > tradeoffs.> The fix would be easy enough -- at a glance there are two usages of > .Date(seq.int(...)) in seq.Date() that could be replaced by > .Date(as.numeric(seq.int(...))).Thanks. Can you pls provide a patch for these? Best -k> Mike C> ______________________________________________ > R-devel at r-project.org mailing list > https://stat.ethz.ch/mailman/listinfo/r-devel