As seen in r-cmd-check-occasional:
|
test(10.16, coerceAs(1:2, factor(c("x","y"))), factor(c("x","y")), output="integer[integer] into integer[factor]") |
When run in a locale where y sorts before x we get:
> x = coerceAs(1:2, factor(c("x", "y")))
First 2 of 2 (type 'integer'):
[1] y x
Levels: y x
> y = factor(c("x", "y"))
First 2 of 2 (type 'integer'):
[1] x y
Levels: y x
2 string mismatches
I am not sure the intended behavior in this case, but that output does look potentially off to me. We can fix the test if the behavior is intentional, want to check with Jan first.
As seen in
r-cmd-check-occasional:data.table/inst/tests/nafill.Rraw
Line 277 in c8c35f3
When run in a locale where
ysorts beforexwe get:I am not sure the intended behavior in this case, but that output does look potentially off to me. We can fix the test if the behavior is intentional, want to check with Jan first.