hi,
if the NaN is the culprit, a way to circumvent the problem could
be using find(isnan(a)) to construct a vector holding the position
of all NaN in your matrix, then replace all NaN with some arbitrary
number before saving the matrix and the NaN-indices vector. Hopefully
you should be able to save the file using the -float-binary option.
You should save the pair matrix and index-vector in the same file to
make sure all information is kept together. Last, you only need to
reverse the process when loading the matrix.
On Tue, Jun 14, 2005 at 04:14:44PM +0200, Søren Hauberg wrote:
a = zeros(3); a(5) = NaN;
octave:25> save -float-binary test.data a
which gives me the following warning:
warning: save: some values too large to save as floats --
warning: save: saving as doubles instead
Btw. I'm using 2.1.71 on linux
Miquel