Re: [sage-devel] Warn about long doctests (#16609 needs review)

What's wrong with "Set plot_points to a higher value to get more detail." On 1 Aug 2014 20:09, "kcrisman" wrote: > 1. Test that code produces the expected output 2. Document what sage can do The 1000 point resolution is not so useful for 1, but is useful for 2 -- it illustrate

Re: [sage-devel] Warn about long doctests (#16609 needs review)

> > 1. Test that code produces the expected output >>> 2. Document what sage can do >>> >>> The 1000 point resolution is not so useful for 1, but is useful for 2 -- >>> it illustrates that sage can produce very high quality graphics >>> >> >> IMHO the doctest should test that the plot_points opti

Re: [sage-devel] Warn about long doctests (#16609 needs review)

On Wednesday, July 30, 2014 5:20:21 PM UTC-4, Volker Braun wrote: > > On Wednesday, July 30, 2014 11:04:02 AM UTC-4, Niles Johnson wrote: >> >> 1. Test that code produces the expected output >> 2. Document what sage can do >> >> The 1000 point resolution is not so useful for 1, but is useful for

Re: [sage-devel] Warn about long doctests (#16609 needs review)

On Wednesday, July 30, 2014 11:04:02 AM UTC-4, Niles Johnson wrote: > > 1. Test that code produces the expected output > 2. Document what sage can do > > The 1000 point resolution is not so useful for 1, but is useful for 2 -- > it illustrates that sage can produce very high quality graphics > IM

Re: [sage-devel] Warn about long doctests (#16609 needs review)

On Tuesday, July 29, 2014 2:24:10 PM UTC-4, Volker Braun wrote: > > > What do we test with 1000 point resolution that can't be tested with 10? > Maybe more examples should be flagged with "# not tested" or something like "# optional -- demo" or "# optional -- tutorial"? This and another recent

Re: [sage-devel] Warn about long doctests (#16609 needs review)

On Tuesday, July 29, 2014 5:27:50 PM UTC+2, Ralf Stephan wrote: > > On Tuesday, July 29, 2014 2:32:43 PM UTC+2, Volker Braun wrote: >> >> (105.58554887771606, u'sage.functions.bessel'), >> > > It's just four complex_plots that need all long time (70% of total) here. > I thought setting plot_point

Re: [sage-devel] Warn about long doctests (#16609 needs review)

On Tuesday, July 29, 2014 8:59:56 PM UTC-4, P Purkayastha wrote: > > Hmm.. sage.plot.{graphics,plot} can not be avoided. We need to show a lot > of examples to the users, since the number of possible customizations is > huge. > And they don't trigger the warnings since each individual tests is

Re: [sage-devel] Warn about long doctests (#16609 needs review)

Hmm.. sage.plot.{graphics,plot} can not be avoided. We need to show a lot of examples to the users, since the number of possible customizations is huge. On Tuesday, July 29, 2014 8:32:43 PM UTC+8, Volker Braun wrote: > > sage: stats = json.load(open('/home/vbraun/.sage/timings2.json')) > sage: s

Re: [sage-devel] Warn about long doctests (#16609 needs review)

On Tuesday, July 29, 2014 4:20:09 AM UTC-4, John Cremona wrote: > > The other factor is that some modules are themselves very long (thousands > of lines) withough -- necessarily -- having long individual tests. Though > probably they do. > Only the time needed for a single test is used to decid

Re: [sage-devel] Warn about long doctests (#16609 needs review)

This is with the libgap saved workspace (and initializing libgap takes only one second on my laptop anyways) On Tuesday, July 29, 2014 1:25:41 PM UTC-4, Travis Scrimshaw wrote: > > Is this with the libgap speedup? I know a bunch of those end up importing > libgap (usually through a Weyl group).

Re: [sage-devel] Warn about long doctests (#16609 needs review)

Is this with the libgap speedup? I know a bunch of those end up importing libgap (usually through a Weyl group). The tests in rigged_configurations are there for coverage (and have been useful for catching bugs). I can look into some of the other combinat ones. Best, Travis On Tuesday, July 2

Re: [sage-devel] Warn about long doctests (#16609 needs review)

On Tuesday, July 29, 2014 2:32:43 PM UTC+2, Volker Braun wrote: > > (105.58554887771606, u'sage.functions.bessel'), > It's just four complex_plots that need all long time (70% of total) here. Regards, -- You received this message because you are subscribed to the Google Groups "sage-devel"

Re: [sage-devel] Warn about long doctests (#16609 needs review)

OK, so I may have some responsibility for 2 of those: (108.91321992874146, u'sage.schemes.elliptic_ curves.isogeny_small_degree'), (203.31193494796753, u'sage.schemes.elliptic_ curves.ell_rational_field'), and it should be easy to reduce the load for the first one. I doubt that nagging will work

Re: [sage-devel] Warn about long doctests (#16609 needs review)

sage: stats = json.load(open('/home/vbraun/.sage/timings2.json')) sage: sorted([(v['walltime'], k) for k,v in stats.items() if 'failed' not in v])[-20:] [(103.71118903160095, u'sage.combinat.rigged_configurations.tensor_product_kr_tableaux'), (105.40626382827759, u'sage.combinat.root_system.plo

Re: [sage-devel] Warn about long doctests (#16609 needs review)

I would not be surprised if some of the offending long tests are by me (" just because the author wants to show off the capabilities of their code") so can you post a list of the longest ones? The other factor is that some modules are themselves very long (thousands of lines) withough -- necessari

[sage-devel] Warn about long doctests (#16609 needs review)

A small handful of modules takes the majority of the doctest walltime, see e.g. this lopsided distribution: I'm proposing to * turn the existing sage -t --warn-long parameter into a warning (instead of an error) * enable it by defau