Hi Florent, On Thu, 24 Nov 2011 22:20:26 +0100 Florent Hivert <florent.hiv...@lri.fr> wrote:
> I'm working on #11576 "make it possible to generate sequences of > variables easily". I actually need to achieve a larger goal: I need > to have symbolic variables indexed by any Sage objects (eg: integers, > group element, matrices...). As as said previously here, I started to > work on Burcin patch. So I extended a little Burcin patch. Now I'm > faced with the problem that GiNaC indexed variable seems to have a > not trivial semantic that I don't understand at all. Moreover I don't > know how to debug Cython / C++ code so I'm quite stuck. Here is the > problem I have: > > To get it you should apply > attachment:indexed_expression-experimental-fh.patch from #11576 on a > fresh Sage-4.7.2 install. Then I can Your changes seem to be based on an older version of my patch. The latest version was here: http://sage.math.washington.edu/home/burcin/indexed_expression_20111118.patch I used a dummy symbolic variable as the dimension when it's not specified. This doesn't make any difference though. > To make thing crystal clear (I hope): > > sage: class bla(SageObject): > ... def __add__(self, other): > ... print "Addition called" <snip> Your example was trying to multiply two instances of the class bla. Here is what I guess you meant to write: sage: m1 = bla() sage: m2 = bla() sage: a, b = x.ind[m1],2*x.ind[m2] Traceback (most recent call last) <snip> in sage.symbolic.expression.Expression._add_ (sage/symbolic/expression.cpp:11082)() TypeError: unsupported operand type(s) for -: 'bla' and 'bla' Note that it is trying to subtract the objects. Probably trying to check equality. :) > WTF ! Why is it adding or multiplying my indices for nothing ! It is > a problem of Ginac ? of the wrapper ? or behind the chair and the > screen ? Here is the backtrace: Breakpoint 2, GiNaC::numeric::sub (this=0x7fffffffb4a0, other=...) at numeric.cpp:1745 1745 { (gdb) call this->dbgprint() <class '__main__.bla'> (gdb) call other.dbgprint() <class '__main__.bla'> (gdb) bt #0 GiNaC::numeric::sub (this=0x7fffffffb4a0, other=...) at numeric.cpp:1745 #1 0x00007fffd687c599 in GiNaC::operator- (lh=<value optimized out>, rh=<value optimized out>) at operators.cpp:97 #2 0x00007fffd6879226 in GiNaC::numeric::compare_same_type (this=0x4796060, other=...) at numeric.cpp:1700 #3 0x00007fffd67ca25a in compare (this=0x3d9cd10, other=...) at ex.h:339 #4 GiNaC::idx::compare_same_type (this=0x3d9cd10, other=...) at idx.cpp:278 #5 0x00007fffd6ce0aad in compare (this=<value optimized out>, other=<value optimized out>) at /home/burcin/sage/sage-4.7.2/local/include/pynac/ex.h:339 #6 GiNaC::container<std::vector>::compare_same_type ( this=<value optimized out>, other=<value optimized out>) at /home/burcin/sage/sage-4.7.2/local/include/pynac/container.h:625 #7 0x00007fffd67b5f5e in compare (this=0x45508e0, lh=<value optimized out>, rh=...) at ex.h:339 #8 GiNaC::expairseq::construct_from_2_ex (this=0x45508e0, lh=<value optimized out>, rh=...) at expairseq.cpp:887 #9 0x00007fffd6771beb in GiNaC::add::add (this=0x45508e0, lh=..., rh=...) at add.cpp:70 #10 0x00007fffd687c46c in exadd (lh=..., rh=...) at operators.cpp:42 #11 GiNaC::operator+ (lh=..., rh=...) at operators.cpp:69 #12 0x00007fffd644e489 in __pyx_f_4sage_8symbolic_10expression_10Expression__add_ (__pyx_v_left=0x4757050, __pyx_v_right=0x4757d40, In English: it tries to normalize the resulting expression in construct_from_2_ex(). This means checking if the two arguments are equal. Going through idx::compare_same type(), it ends up in numeric::compare_same_type(). Here is the code for that function: int numeric::compare_same_type(const basic &other) const { GINAC_ASSERT(is_exactly_a<numeric>(other)); const numeric &o = static_cast<const numeric &>(other); int cmpval = (real() - o.real()).csgn(); if (cmpval != 0) return cmpval; return (imag() - o.imag()).csgn(); } We should change this somehow. Maybe we could try PyObject_Compare and fall back to this if it fails. Comments? Cheers, Burcin -- To post to this group, send an email to sage-devel@googlegroups.com To unsubscribe from this group, send an email to sage-devel+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/sage-devel URL: http://www.sagemath.org
