You are right in that most of the time when I start to expand the window I
will get the right image. However sometimes even with that action  I get the
wrong image ( vague , hazy lines as opposed to squares. Also in ijx with
debug on which I used to stop my verb before the savemat action I get the
same result in the ijx window.
My images are simply like density plots and I will look at image3 addon. Is
it a matter of replacing 2 lines with 2 lines in my verb or something much
more complex ( nothing to do with complex numbers) than that.

cheers


Roelof

"Have a cool mind and a warm heart" - Ajahn Sona
 "Like ......(hamsters on a wheel), (ambitious people) still climb and
climb, with great labor, and incessant anxiety, but never reach the top."  
Robert Browning
"We shall not cease from exploration. And at the end of all our exploring
Will be to arrive were we started and know the place for the first time" -
T. S. Eliot


Dr. Roelof K. Brouwer PH.D.  P.Eng.
Professor Emeritus
Dept. of Computing Science
School of Advanced Technologies and Mathematics
Thompson Rivers University
900 McGill Road
Kamloops, BC
Canada, V2C 5N3

email : [email protected]
www.tru.ca/advtech/compsci/faculty/rbrouwer


-----Original Message-----
From: [email protected]
[mailto:[email protected]] On Behalf Of Devon McCormick
Sent: December-14-08 9:32 PM
To: Programming forum
Subject: Re: [Jprogramming] savemat_jviewmat_

Dr. Brouwer - I think this is not so much a Vista 64-bit problem but a
general problem with the event-loop/message-pump handling.

I've encountered this problem before: I have a sequence of lines of code
that work fine in immediate execution but fail when they're in a function.
I believe the problem is that viewmat needs to get to a state of waiting for
user input before it will finish displaying its picture.  This doesn't
happen within multiple lines in a function, so only a blank picture is
saved.

I've tried lots of different ways to get around this and have settled on
using the "image3" add-on to write my images instead of relying on viewmat.
This is too bad because viewmat does offer a nice display of complex numbers
that I haven't replicated with image3 code.

This problem is also mentioned by Dan Bron at
http://www.jsoftware.com/jwiki/DanBron/Snippets/PrettyPictures and even Oleg
doesn't have a solution.

Regards,

Devon**

On Sun, Dec 14, 2008 at 12:54 PM, Roelof K Brouwer
<[email protected]>wrote:

> I have problems with viewmat save in vista 64bit. In particular the
> following two statements
>
>
>
>
>
> ( 1 4 14 {". COLOR16) viewmat i.4 4
>
> savemat_jviewmat_ 'C:\Temp\garbage.bmp'
>
>
>
>
>
>  this will give a blank window when used  in a verb
>
> and also in ijx when debug is in
>
> sometimes by adjusting the viewmat window I will get the proper display .
> i.e. click the margin of the window for exdpansion. However sometimes on
> expansion the window does not have the correct
>
> displayI
>
> if I remove the save command things will work
>
>
>
> when in ijx I can first run viewmat command and then the save command and
> things will work
>
> I also tried puttting a delay between the two viewmat commands
>
>
>
> here is my complete file with verb
>
>
>
> drivedata=: 'C:\Users\RKBrouwer\Documents\ACA\Research\ResData\'
>
> drive_op=:'C:\Users\RKBrouwer\Documents\Temp\Set\'
>
> drive=:'C:\Users\RKBrouwer\Documents\ACA\'
>
> require 'viewmat color16 plot'
>
> load 'C:\Users\RKBrouwer\Documents\J_Code_Roelof\J_utilities_roelof.ijs'
>
> NB. load
> 'C:\Users\RKBrouwer\Documents\J_utilities_roelof\visual_clustering.ijs'
>
> require 'jfiles'
>
>
>
> test_viewmat=:3 : 0
>
>                ( 1 4 14 {". COLOR16) viewmat i.4 4
>
>                savemat_jviewmat_ 'C:\Temp\garbage.bmp'
>
> )
>
>
>
>
>
>
>
> cheers
>
>
>
>
>
> Roelof
>
>
>
> "Have a cool mind and a warm heart" - Ajahn Sona
>
>  "Like ......(hamsters on a wheel), (ambitious people) still climb and
> climb, with great labor, and incessant anxiety, but never reach the top."
> Robert Browning
>
> "We shall not cease from exploration. And at the end of all our exploring
>
> Will be to arrive were we started and know the place for the first time" -
> T. S. Eliot
>
>
>
>
>
> Dr. Roelof K. Brouwer PH.D.  P.Eng.
>
> Professor Emeritus
>
> Dept. of Computing Science
>
> School of Advanced Technologies and Mathematics
>
> Thompson Rivers University
>
> 900 McGill Road
>
> Kamloops, BC
>
> Canada, V2C 5N3
>
>
>
> email :  <mailto:[email protected]> [email protected]
>
>  <http://www.tru.ca/advtech/compsci/faculty/rbrouwer>
> www.tru.ca/advtech/compsci/faculty/rbrouwer
>
>
>
> ----------------------------------------------------------------------
> For information about J forums see http://www.jsoftware.com/forums.htm
>



-- 
Devon McCormick, CFA
^me^ at acm.
org is my
preferred e-mail
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

Reply via email to