Repository navigation
Increase tolerances in substract_vec_mat test - #1483
Conversation
|
(stat_comp_benchmarks/benchmarks/gp_pois_regr/gp_pois_regr.stan, 0.99) |
|
I'm guessing it's a different compiler or settings or different hardware.
… On Dec 4, 2019, at 3:55 AM, Rok Češnovar ***@***.***> wrote:
Summary
We are still experiencing some issues on develop after #1469 was merged. This increases the tolerances for another test. If new test failures come up then I would say we should increase the tolerance globally. I will also investigate what causes the "randomness" of this tests that does not use any randomness whatsoever. Might be limited to Windows.
Tests
This increases the tolerances in the tests.
Side Effects
/
Checklist
• Math issue #1453
• Copyright holder: Rok Češnovar
• the basic tests are passing
• the code is written in idiomatic C++ and changes are documented in the doxygen
• the new changes are tested
You can view, comment on, or merge this pull request online at:
#1483
Commit Summary
• increase tolerances in substract_vec_mat test
File Changes
• M test/unit/math/mix/mat/fun/subtract_test.cpp (7)
Patch Links:
• https://github.com/stan-dev/math/pull/1483.patch
• https://github.com/stan-dev/math/pull/1483.diff
—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub, or unsubscribe.
|
|
This is ready for review. I will examine the strange windows behavior in tests in a separate issue/PR. This should be merged to prevent Jenkins issues for other developers. |
4 participants
Summary
We are still experiencing some issues on develop after #1469 was merged. This increases the tolerances for another test. If new test failures come up then I would say we should increase the tolerance globally. I will also investigate what causes the "randomness" of this tests that does not use any randomness whatsoever. Might be limited to Windows.
Tests
This increases the tolerances in the tests.
Side Effects
/
Checklist
Math issue MathMixMatFun.add and MathMixMatFun.subtract tests seem to fail randomly? #1453
Copyright holder: Rok Češnovar
the basic tests are passing
the code is written in idiomatic C++ and changes are documented in the doxygen
the new changes are tested