Bug 102492 - Matrix values are all zero
Summary: Matrix values are all zero
Status: RESOLVED FIXED
Alias: None
Product: kst
Classification: Applications
Component: general (show other bugs)
Version: 1.x
Platform: Compiled Sources Linux
: NOR normal
Target Milestone: ---
Assignee: kst
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2005-03-25 19:57 UTC by Andrew Walker
Modified: 2005-04-04 07:54 UTC (History)
0 users

See Also:
Latest Commit:
Version Fixed In:
Sentry Crash Report:


Attachments

Note You need to log in before you can comment on or make changes to this bug.
Description Andrew Walker 2005-03-25 19:57:53 UTC
Version:           HEAD (using KDE KDE 3.3.0)
Installed from:    Compiled From Sources
OS:                Linux

PROBLEM:
When creating a matrix there are no values associated with it (i.e. the matrix has zero size in the x-direction).

STEPS TO REPRODUCE:
Start Kst
Create a vector
Create a matrix from the created vector
Create an image from the created matrix.

RESULTS:
Nothing is displayed as the matrix has no size.

EXPECTED RESULTS:
The matrix is properly created.
Comment 1 George Staikos 2005-03-28 18:27:29 UTC
Fixed for now, more proper fix to come later
Comment 2 Andrew Walker 2005-03-28 20:38:19 UTC
The exact same problem can still be produced by simply not including the last step of the "steps to reproduce" i.e. don't create the image.

If you create a matrix, and then open the dialog to create an image, select the matrix recently created and hit Auto Threshold the limits produced are incorrect, as the matrix has no values. Also, viewing the matrix values (Data... View Matrix Values) shows that the matrix is empty.
Comment 3 George Staikos 2005-03-28 20:42:07 UTC
On Monday 28 March 2005 13:38, Andrew Walker wrote:
> ------- The exact same problem can still be produced by simply not
> including the last step of the "steps to reproduce" i.e. don't create the
> image.
>
> If you create a matrix, and then open the dialog to create an image, select
> the matrix recently created and hit Auto Threshold the limits produced are
> incorrect, as the matrix has no values. Also, viewing the matrix values
> (Data... View Matrix Values) shows that the matrix is empty.


  That's by design.  It's unused so no need to update and waste cpu cycles.
Comment 4 Andrew Walker 2005-03-28 20:55:30 UTC
Sadly this design makes auto-threshold useless, but auto-threshold is necessary as the user may not know the range of their data. I would propose that we revert to the original design.
Comment 5 Netterfield 2005-03-28 21:00:14 UTC
On March 28, 2005 01:42 pm, George Staikos wrote:
> On Monday 28 March 2005 13:38, Andrew Walker wrote:
> > ------- The exact same problem can still be produced by simply not
> > including the last step of the "steps to reproduce" i.e. don't create the
> > image.
> >
> > If you create a matrix, and then open the dialog to create an image,
> > select the matrix recently created and hit Auto Threshold the limits
> > produced are incorrect, as the matrix has no values. Also, viewing the
> > matrix values (Data... View Matrix Values) shows that the matrix is
> > empty.
>
>   That's by design.  It's unused so no need to update and waste cpu cycles.


With the new design, clearly auto threshold and view values need to be 
preceeded by an update on objects being viewed and autothresholded...

Same as scalar accesses.  George, is this in the works?
Comment 6 George Staikos 2005-03-28 21:12:33 UTC
On Monday 28 March 2005 13:55, Andrew Walker wrote:
> ------- Sadly this design makes auto-threshold useless, but auto-threshold
> is necessary as the user may not know the range of their data. I would
> propose that we revert to the original design.


   Let's slow down and take this one step at a time.  It's quite possible that 
the matrix and image code has to be redesigned.  I'm already working on part 
of it.
Comment 7 George Staikos 2005-03-28 21:20:48 UTC
On Monday 28 March 2005 14:00, netterfield@astro.utoronto.ca wrote:
> With the new design, clearly auto threshold and view values need to be
> preceeded by an update on objects being viewed and autothresholded...
>
> Same as scalar accesses.  George, is this in the works?


   No plan for it yet.  I wasn't aware of this stuff.  However the updates 
-have- to happen in the update thread.  The first answer that comes to mind 
is that anything that depends on the content of an object has to check the 
dirty flag and delay its action until the object is no longer dirty.  Just 
like with vectors, there is no guarantee of any content at the start.
Comment 8 George Staikos 2005-04-04 07:54:51 UTC
*** Bug has been marked as fixed ***.