octave-maintainers
[
Top
][
All Lists
]
Advanced
[
Date Prev
][
Date Next
][
Thread Prev
][
Thread Next
][
Date Index
][
Thread Index
]
RE: Sparse matrix problem
From
:
michael . goffioul
Subject
:
RE: Sparse matrix problem
Date
:
Wed, 20 Dec 2006 11:26:04 +0100
I was wrong when I stated it worked OK for full matrices and not
for sparse matrices:
[1+i, 2-i] ./ [0+1i, 0] => [1-1i, NaN+NaNi]
sparse([1+i, 2-i]) ./ sparse([0+1i, 0]) => [1-1i, NaN+NaNi]
Actually, this is due to the complex class implementation. From simple
test, it appears that dividing by std::complex<double>(0.0,0.0) leads
to (NaN,NaN). When dividing a complex object by 0.0, it leads to
(Inf,Inf). So this is not octave's fault.
Michael.
reply via email to
[
Prev in Thread
]
Current Thread
[Next in Thread]
Sparse matrix problem
,
michael . goffioul
,
2006/12/19
Re: Sparse matrix problem
,
David Bateman
,
2006/12/19
RE: Re: Sparse matrix problem
,
michael . goffioul
,
2006/12/19
RE: Sparse matrix problem
,
michael . goffioul
<=
Prev by Date:
RE: Re: Sparse matrix problem
Next by Date:
MSVC patch: export symbols in macros
Previous by thread:
RE: Re: Sparse matrix problem
Next by thread:
MSVC patch: export symbols in macros
Index(es):
Date
Thread