William Stein wrote: > On Mon, Mar 24, 2008 at 12:53 PM, Jason Grout > <[EMAIL PROTECTED]> wrote: >> >> >> When applying the patch for a overhauled matrix() function at >> http://trac.sagemath.org/sage_trac/ticket/2651, I get doctest failures >> in crypto/mq/sr.py and sage/plot/plot3d/transform.pyx related to calling >> matrix() with a list of rows, but specifying a number of rows that >> conflicts. >> >> You can see these failures by applying the patch and running sage -t >> -long on either devel/sage/sage/crypto/mq/sr.py or >> devel/sage/sage/plot/plot3d/shapes2.py (and the same failures make a >> whole bunch of other doctests fail too). >> >> More specifically, for sr.py, matrix() is called from lines 1775 and >> 1779 with a list of lists. The code appears to flatten "l" into a flat >> list, but the flatten command on the preceding lines specifies to only >> flatten Vector_modn_dense things, so the list "l" is not flattened. >> >> For transform.pyx, the call to matrix on line 44 appears to flatten the >> trans argument (i.e., list(trans)), but many times what is actually >> passed to Sage is a list containing a single RDF vector instead. >> >> I'm not familiar with the code base for these two files enough to >> propose fixes. One frustrating thing is that when I run the doctests by >> hand, everything seems to work fine. I'm not sure what the difference >> is between running the tests by hand or using sage -t -long. >> >> Can someone take a look at one or both of these bugs? If we can resolve >> these, I think the matrix() rewrite at #2651 is ready to be reviewed >> (after one last small cleanup). >> > > It sounds to me like the matrix rewrite should be considered done, then > two new tickets should be opened for the above two issues. Martin > Albrecht write sr.py and Robert Bradshaw wrote transform.pyx. >\
malb posted a patch which fixes the sr.py issue on the ticket (#2651). I've updated the patch on the ticket and it's now ready to be reviewed. The only tests I can see failing (that are related on 2.11alpha1) are the tests which fail because of the error in transform.pyx noted above. The patch also fixes #2658. The transform.pyx bug is now #2667. Jason --~--~---------~--~----~------------~-------~--~----~ 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 -~----------~----~----~----~------~----~------~--~---
