Bug 744 - The umeyama_5 test fails for seed 1392390306
The umeyama_5 test fails for seed 1392390306
Status: RESOLVED FIXED
Product: Eigen
Classification: Unclassified
Component: Geometry
3.2
All All
: Normal Unknown
Assigned To: Nobody
:
Depends on:
Blocks:
  Show dependency treegraph
 
Reported: 2014-02-14 19:10 UTC by Benoit Steiner
Modified: 2014-02-17 13:55 UTC (History)
4 users (show)



Attachments

Description Benoit Steiner 2014-02-14 19:10:01 UTC
I compiled the latest version of the Eigen codebase for x64 with SSE3 enabled. ./umeyama_5 s1392390306 fails as follow:
Initializing random number generator with seed 1392390306
Repeating each test 10 times
Test (run_fixed_size_test<float, 4>(num_elements)) failed in /usr/local/google/home/bsteiner/EigenLatest/eigen/test/umeyama.cpp (154)
    error < Scalar(10)*std::numeric_limits<Scalar>::epsilon()
Stack:
  - (run_fixed_size_test<float, 4>(num_elements))
  - umeyama
  - Seed: 1392390306

Aborted (core dumped)
Comment 1 Hauke Heibel 2014-02-16 19:55:21 UTC
I just tried this on Visual Studio 2013 and cannot reproduce the issue. Which compiler did you use?
Comment 2 Benoit Steiner 2014-02-17 02:23:37 UTC
I used gcc4.8 running on Ubuntu.
Comment 3 Christoph Hertzberg 2014-02-17 09:42:10 UTC
I can confirm this on Ubuntu with gcc4.6.3. Moreover, with 10000 repetitions it fails almost every time -- with 1000 repetitions ca 1 out of 5 times.
I assume these are merely false negatives and we should just slightly reduce the precision (maybe 16*eps instead of 10*eps).
Comment 4 Christoph Hertzberg 2014-02-17 13:55:10 UTC
Another problem was that the problem got ill-posed if the scaling factor was very small compared to the translation part.
I relaxed the test-case here:
https://bitbucket.org/eigen/eigen/commits/50e33ba
https://bitbucket.org/eigen/eigen/commits/3e5302f

Note You need to log in before you can comment on or make changes to this bug.