#14284: Sampling in unit tests
-------------------------------+--------------------------------------------
Reporter: roed | Owner: jason
Type: enhancement | Status: needs_work
Priority: major | Milestone: sage-5.9
Component: misc | Resolution:
Keywords: | Work issues:
Report Upstream: N/A | Reviewers: Julian Rueth
Authors: David Roe | Merged in:
Dependencies: #14285 | Stopgaps:
-------------------------------+--------------------------------------------
Changes (by roed):
* status: positive_review => needs_work
Comment:
Replying to [comment:9 nthiery]:
> Hi!
>
> Sorry to jump in a bit late in the discussion; I had not noticed this
ticket before.
Well, we did just open it a couple days ago (Julian's visiting me in
Calgary so we're working in person).
> I definitely see and approve the point of the ticket. On the other hand,
I find the current idiom to be used in {{{_test_associativity}}} and
friends a bit heavy. What about an idiom like:
>
> {{{
> S = tester.some_elements()
> for x,y,z in tester.some_elements(CartesianProduct(S,S,S)):
> ...
> }}}
>
> It's short, and encapsulate TestSuite's inner logic for testing
strategies (on how many elements to run the tests, whether to do tests at
random or not, ...).
>
> This of course requires implementing:
> {{{
> tester.some_elements(XXX)
> }}}
>
> The default implementation could be to run XXX.some_elements(). Or
iterate through XXX, stopping if there are more than n_max elements. Or
take a sample if XXX implements sample. Or ...
>
> What do you think? If you agree, then I would suggest doing the changes
in this ticket, in order to minimize changes and counter changes (they
probably will require some rebasing of my upcoming category patches; the
less rebasing the better :-)).
Sounds okay to me: I'll make a new version. There will be a couple tests
that can't use this idiom (`_test_eq_symmetric` for example), but they can
certainly use the current approach.
>
> By the way: for reproducibility of test failures, I am not super
comfortable with random testing as a default; but that might be just me.
Since Sage sets the random seed at the beginning of each run, tests that
use randomness should be reproducible....
--
Ticket URL: <http://trac.sagemath.org/sage_trac/ticket/14284#comment:10>
Sage <http://www.sagemath.org>
Sage: Creating a Viable Open Source Alternative to Magma, Maple, Mathematica,
and MATLAB
--
You received this message because you are subscribed to the Google Groups
"sage-trac" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To post to this group, send email to [email protected].
Visit this group at http://groups.google.com/group/sage-trac?hl=en.
For more options, visit https://groups.google.com/groups/opt_out.
