MatView
MatView: Scalable Sparse Matrix Viewer
MatView is a handy tool for viewing and exploring large sparse matrices.
Using MatView, a sparse matrix with millions of elements can be reduced
to a graphically viewable size
for overviews of the high-level structure,
or a detailed view can be obtained by zooming in and navigating
through small regions of the matrix.
MatView provides:
- Scalability: Efficient Viewing of Very Large Matrices
(including new huge matrix file support!).
- Easy Thumbnail Navigation, and Zooming In & Out of Matrix Details.
- A Variety of Data Reduction Functions
(Avg, Min/Max, Sum, Std, Var).
- Custom Color Adjustments, or Greyscale Rendering.
- Viewing of Matrix Pattern Only, If Desired.
- Filtering of Matrix Values Outside the Color Map Range, if desired.
- Printing of Matrix Views, or Saving Views as PostScript / PPM.
- Compatibility with Matrix Market Format, Harwell-Boeing Format,
and Matlab MAT Files.
- Loading of Compressed or Gzipped Matrix Files.
The MatView 1.0 Beta 6 Software
(no longer Alpha, complete with new "configure" installation!)
is available for
FREE DOWNLOAD (gzipped tar format -
you can also download
uuencoded
and
zipped
distributions).
Now Works with Tcl/Tk 8.4! :-D (12/3/2008)
(Now with smaller example matrices! :-)
Interface Samples:
Main Interface:
Control Panel:
Separate Viewing Window:
Color Mapping Function Warping:
Project Status:
MatView is built using the Tcl/Tk
system, and so should theoretically run anywhere that Tcl/Tk runs...
(so far tested only on Unix platforms).
MatView is an ongoing project, with ever-continuing user
interface additions and improvements, and there are plans to add
pluggable modules for simple matrix transformations,
matrix graphing support, and support for more matrix file formats.
This work is preliminary and experimental,
and is intended only as an assistant to
more elaborate linear algebra solvers and systems.
Please feel free to download MatView, give it a try, and send me
back any feedback regarding its usefulness (or lack thereof :-).
For more details on MatView or any questions, email Jim Kohl:
kohlja at ornl dot gov
Distributed Computing Group
Computer Science and Mathematics Division
Oak Ridge National Laboratory
P.O. Box 2008, Bldg 6012, MS 6367
Oak Ridge, TN 37831-6367
Research supported by the Applied Mathematical Sciences
Research Program, Office of Energy Research, U. S. Department of Energy,
under contract No. DE-AC05-00OR22725 with UT-Battelle, LLC.
http://www.csm.ornl.gov/~kohl/MatView/
Last modified: December 3, 2008 by
Kohl.