On Jan 17, 2008 8:15 AM, mabshoff <[EMAIL PROTECTED]> wrote: > > > > On Jan 17, 11:39 am, Robert Bradshaw <[EMAIL PROTECTED]> > wrote: > > Hello, > > > transform.pyx does some maxima stuff, perhaps the patch (or some > > combination with other patches) is leaving maxima in an unstable state? > > I consider this the most likely culprit, but I am not familiar with > the maxima interface, so somebody else needs to debug this. I can't > wait for sympy to step up and at least do symbolic matrices. Ondrej? > > > > - Robert > > > > On Jan 17, 2008, at 1:46 AM, Simon King wrote: > > > > > Hi Michael, > > > > > since i was partially involved in that patch, i'd like to know what > > > happened. > > > > > On Jan 17, 4:32 am, mabshoff <[EMAIL PROTECTED] > > > dortmund.de> wrote: > > >> I am as surprised as you are that this patch causes problems. What > > >> happens is that when doctesting plot/plot3d/transform.pyx it just > > >> hangs. > > > > > What does transform.pyx do? Does it hang since it computes some > > > modular invariant ring? This might be not a surprize: Singular is much > > > faster than Magma in computing *non*-modular invariant rings, but > > > Magma is much faster in *modular* invariant rings. > > > > > Or is plot/plot3d/transform.pyx doing something completely different, > > > so that it seems there is an odd side-effect? > > > > >> The patch itself did get applied correctly and I don't > > >> see how it could interfere, but it does. Since maxima is involved in > > >> this doctest I will leave this for somebody else who is more familiar > > >> with that subsystem to debug. > > > > > Perhaps this is related to something that i met a while ago. I had > > > some computation involving occasionally some logarithms. When i > > > repeated the *same* computation very often, the computation time for > > > each run dramatically increased. That's to say: If t is the time for > > > the first run of the computation, then the time for run number 100 has > > > been 20*t or so. A closer inspection showed that this increase only > > > concerned Maxima computations. I'll try to reproduce that phenomenon. > > I remember the issue, but the doctest just sits there at 0% CPU for > hours, so I don't think it is that problem. Most likely as Robert > suggested above it is a synchronization issue with the Maxima > interface. > > The interesting question: Can anybody reproduce this with 2.10.alpha4 > + the bundle from #1485 applied?
Sorry for the long delay - it took me awhile to try this. 1) I installed 2.10.alpha4 and created a clone 2) I simply copied http://sage.math.washington.edu/home/wdj/patches/matrix_group.py to the appropriate place (couldn't figure out how to apply the patch...) 3) rebuilt sage using sage -b (twice) 4) ran sage-testall All tests passed. > > > > Yours > > > Simon > > Cheers, > > Michael > > > > --~--~---------~--~----~------------~-------~--~----~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://www.sagemath.org -~----------~----~----~----~------~----~------~--~---
