Hi,
Indeed that is a problem with all interfaces that are text
based such as magma (to communicate with Magma Sage uses
an emulation of a console). You can analyze what is sent
to magma by running
sage: m = matrix(2, [1,3,2,4])
sage: m._magma_init_(magma)
'Matrix(IntegerRing(),2,2,StringToIntegerSequence("1 3 2 4"))'
As I don't have magma I can not test it for sparse matrices,
but you can have a look at
sage: m = matrix(2, [0,0,2,0], sparse=True)
sage: m._magma_init_(magma)
For sparse matrices, it might be the case that the
string produced is somehow too big for no good reason. If
you find a smarter way to pass the data that would speed
up the process and could be incorporated into Sage.
Best
Vincent
Le 26/02/2019 à 17:36, Pierre Guillot a écrit :
Hi !
I'm trying to compute the determinant of a sparse, square matrix with
50,000 columns and coefficients in GF(5)["x y z t u"]. I'm trying my luck
with the magma interface, since we have a copy of magma running at my
university.
The first thing surprising me is that, on toy examples, magma seems pretty
slow, but that may be for another question. At least it works.
But when I try
foo= magma(my_matrix)
with the real example, well I seem to have to wait forever. By contrast,
when the matrix was saved as a .sobj file and then re-loaded back, it tooks
less than 10 seconds. Why should it be so hard to enter it into Magma?
is there any workaround that is worth trying? I could somehow directly
write magma code into a file, and then hope that loading that would be
faster, but I'm not sure why i would do better than the existing magma
interface...
thanks!
Pierre
--
You received this message because you are subscribed to the Google Groups
"sage-support" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to sage-support+unsubscr...@googlegroups.com.
To post to this group, send email to sage-support@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-support.
For more options, visit https://groups.google.com/d/optout.
