OK, this’ll be a bit vague, as I’m working on some unpublished code, but maybe someone can still spot what’s going on…
I have a
testset that looks something like thus:
@testset begin x = Something() # various functions with x as parameter # various tests end
(The code involves building a mixed-integer program based in part on
x, and then solving it using
Now I’d like to use the same test with more types for
x. I started with a minor refactor:
@testset begin for x in [Something()] # various functions with x as parameter # various tests end end
I thought this ought to be about as fast (until I added other values – especially of other types), but for some reason it went from about 0.12 s to 0.27 s.
I also tried swapping the
end of the
testset with an equivalent loop, but that ended up at about 0.30 s.
There’s no random test data generated or anything, which could have caused the model to be harder to solve for
Cbc, for example, as far as I can see.
This is quite slow (surprisingly so, even) to begin with, and so I’d rather avoid doubling it :-} Any thoughts – if nothing else, then on how to debug/figure out what’s going on?
(Using 1.5.0-rc1.0 on macOS Catalina.)