Okay, I got some new results:
Changing permanent() slightly in matrix2.pyx:
print "entering permanent() - m: ",m," n: ",n
from sage.rings.arith import binomial
for r from 1 <= r < m+1:
lst = _choose(n, r)
print "lst", lst
tmp = []
for cols in lst:
tmp.append(self.prod_of_row_sums(cols))
s = sum(tmp)
# sn = (-1)^(m-r)
if (m - r) % 2 == 0:
sn = 1
else:
sn = -1
perm = perm + sn * binomial(n-r, m-r) * s
print "exiting permanent() - perm: ", perm
Where _choose is defined as:
def _choose(int n, int t):
"""
Returns all possible sublists of length t from range(n)
Based on algoritm L from Knuth's taocp part 4: 7.2.1.3 p.4
AUTHOR:
-- Jaap Spies (2007-10-22)
"""
x = []
c = range(t)
c.append(n)
c.append(0)
j = 0
while j < t:
x.append(c[:t])
j = 0
while c[j]+1 == c[j+1]:
c[j] = j
j = j+1
c[j] = c[j]+1
return x
[end of _choose code]
At some point the refcount for the list elements goes FUBAR (this is
not the first occurence of this, but those scrolled off my screen):
We should get (r=1 in this case):
entering permanent() - m: 7 n: 7
lst [[1],[2],[3],[4],[5],[6]]
But we get:
entering permanent() - m: 7 n: 7
lst [[<refcnt -2137432134 at 0x92f70ac>], [<refcnt -2137432183 at
0x92f70a0>], [<refcnt -2137464436 at 0x92f7094>], [<refcnt -2138866030
at 0x92f7088>], [4], [5], [6]]
I assume on 64 bit boxen the refcount is 64 bit, so we never see the
issue on sage.math.
So does anybody have any idea what is wrong?
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://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/
-~----------~----~----~----~------~----~------~--~---
