[SciPy-User] number of tests

[Davide Del Vento]

I should have added:

$ lsb_release -a
LSB Version: 
:core-4.0-amd64:core-4.0-noarch:graphics-4.0-amd64:graphics-4.0-noarch:printing-4.0-amd64:printing-4.0-noarch
Distributor ID: RedHatEnterpriseServer
Description:    Red Hat Enterprise Linux Server release 6.2 (Santiago)
Release:        6.2
Codename:       Santiago

$ python -c "import scipy; scipy.test('full')"
Running unit tests for scipy
NumPy version 1.6.2
NumPy is installed in 
/opt/numpy/1.6.2/intel/13.0.1/lib/python2.7/site-packages/numpy
SciPy version 0.11.0
SciPy is installed in 
/opt/scipy/0.11.0/intel/13.0.1/lib/python2.7/site-packages/scipy
Python version 2.7.3 (default, Feb  9 2013, 16:14:16) [GCC 4.7.2]
nose version 1.2.1

Does anybody knows why the number of tests run are different among 
different runs of the same binary/library on different nodes?
https://github.com/numpy/numpy/blob/master/doc/TESTS.rst.txt implies 
they shouldn't...

Thanks!!

Davide Del Vento,


On 02/11/2013 09:00 PM, Davide Del Vento wrote:
> I compiled scipy 0.11.0 myself with the intel compiler, on top of
> numpy 1.6.2 (with intel compiler too and MKL). I'm trying to assess
> whether or not everything has been build fine. Since my machine is
> actually a cluster, I'm running the tests in different configurations
> (login node and batch script). However, I'm confused by the number of
> tests which ran.
> On the login nodes (either interactively or in a script without the
> tty) I get:
>
> Ran 6218 tests in 514.867s
> FAILED (KNOWNFAIL=17, SKIP=42, errors=1, failures=1)
>
> Whereas in a remote batch node (with a script) I get:
>
> Ran 6167 tests in 271.833s
> FAILED (KNOWNFAIL=17, SKIP=42, errors=201, failures=1)
>
> I am not worried about the difference in timing, since it's about what
> I expected. However, I am surprised about the different number of
> tests that ran, because I'd expect it to be the same (possibly with
> difference in success vs failure vs skipped, but not in the overall
> number).
> Instead I've got 51 "missing" test while running in a remote node.
> Should have been the number of errors 52, I'd suspect an odd way of
> counting what had run, however the difference in errors is 200, so it must
> be something else.
>
> Why is it so?
>
> Thanks and Regards,
> Davide