Directly support array1d format in MatrixMarket I/O

[GoogleCodeExporter]

Make it possible to 
  1) read an Nx1 sparse or dense matrix into an array1d
  2) write an array1d into a Nx1 dense matrix

Original issue reported on code.google.com by wnbell on 5 May 2011 at 3:52

[GoogleCodeExporter]

Here's a patch which does pretty much the same things as posted on the mailing 
list.
I didn't commit it because of the pending release.
I'm also considering the possibility of introducing array1d<->matrix 
conversions (which would be an alternative solution).
What do you think?

Original comment by filipe.c...@gmail.com on 6 May 2011 at 7:07

Attachments:

• mm_vector.diff

[GoogleCodeExporter]

Hi Filipe,

The patch looks good to me, so I think it's safe to integrate now.

array1d<->matrix conversions would also be useful.  I think all you'd need to 
do is introduce an array1d<->array2d conversion and then provide an array1d<->X 
case that used the array1d<->array2d<->X conversion for all the sparse formats.

Original comment by wnbell on 6 May 2011 at 3:11

• Added labels: Milestone-Release0.2
• Removed labels: Milestone-Release0.3

[GoogleCodeExporter]

This issue was closed by revision 2f24c49140be.

Original comment by filipe.c...@gmail.com on 9 May 2011 at 1:20

• Changed state: Fixed