foldl-statistics-0.1.4.1 LinRegState Semigroup testcase

[juhp]

In Stackage Nightly one testcase failed:

      LinRegState Semigroup:                 FAIL (0.19s)
        *** Failed! Falsifiable (after 28 tests and 23 shrinks): 
        [(-40.96317088163834,0.0),(-23.09878837980997,0.0),(-22.00861641446635,1.0)]
        [(10.277689040163754,0.0),(-32.416612824084986,0.0),(-120.15609657741514,0.822813069381938),(18.047491675541508,0.0),(-5.443624038374302,0.0),(-12.563647685049569,0.0),(-86.81658742129697,0.0)]
        Use --quickcheck-replay '27 TFGenR A822A66AAE1CD18840CD7862FAE2DDEAFF10C72ABECF5388BCE8101E9E9266DB 0 1023 10 0' to reproduce.

1 out of 24 tests failed (0.95s)

[axman6]

This is probably due to an overly optimistic test case, it succeeds more often than it fails, but the failure afaict is basically noise, one bit going one way or another depending on order of operations. (you might notice I ran into this too and annotated the tests with a failing case too). I'll look into it, though and see if it's actually a problem. Rerunning the build on Travis more often than not causes the test to pass again.

[juhp]

Okay thanks - I will add it to expected failures in Stackage Nightly then.