Author Topic: qview data table (OPEN M0818)  (Read 733 times)

michaelaye

  • Ra (Power Member)
  • *****
  • Posts: 138
qview data table (OPEN M0818)
« on: January 27, 2012, 05:23:02 PM »
Hi ISIS team,

I would like to ask if it's normal to have the attached appearance happening when accessing the data table.
What do I do?
I link 6 geo-referenced CTX images with each other using the Find tool.
I then use the spatial plot tool to draw a line across the points I am interested to have a profile of.
This creates a nice plot for which I want to have the data.
Now, even so all lines for all 6 images where landing inside their data frames, so no off pixels should be selected (see screenshot), the table starts with: pixel 1-9, then pixel numbers higher than the last pixel at the end of the table, then a full sequence of pixel numbers starting at 1, raising until the end of the table (see 2nd attached screenshot).

Is that maybe a bug of the new plotting system? Or don't I understand something? (I am new to plotting/profiling with qview)

Best regards and a nice weekend,
Michael
« Last Edit: January 31, 2013, 07:53:31 PM by tbecker »

janderson

  • Isis Support Team
  • Isis (Extreme Power Member)
  • *****
  • Posts: 1041
    • http://isis.astrogeology.usgs.gov
Re: qview data table
« Reply #1 on: February 11, 2012, 03:50:02 PM »
Hi Michael,

This looks like a bug to me when you have multiple images linked and plot.

Does the plot tool work okay for you on a single image?

Jeff

michaelaye

  • Ra (Power Member)
  • *****
  • Posts: 138
Re: qview data table
« Reply #2 on: February 14, 2012, 03:06:12 PM »
I'm afraid, it's even harder to reproduce, because sometimes (I'd say 50/50 currently) it does not give lines with N/A with multiple files (I tried up to 12), so it CAN work fine with multiple files as well.

michaelaye

  • Ra (Power Member)
  • *****
  • Posts: 138
Re: qview data table
« Reply #3 on: February 21, 2012, 06:11:54 AM »
qview crashed on me (report attached), this crash might or might not be related:
Before it crashed I noticed that I could not record points anymore, the table display seemed to have added lines but they were not being displayed.

Best regards,
Michael

ssides

  • Isis Support Team
  • Isis (Extreme Power Member)
  • *****
  • Posts: 660
    • http://isis.astrogeology.usgs.gov
Re: qview data table (OPEN INTERNAL)
« Reply #4 on: May 02, 2012, 10:27:56 AM »
Michael,

This issue has been moved to our internal bug tracking system as issue #818

lweller

  • Isis Support Team
  • Isis (Extreme Power Member)
  • *****
  • Posts: 349
Re: qview data table (OPEN M0818)
« Reply #5 on: January 31, 2013, 04:00:57 PM »
Hi Michael,

I have been able to consistently reproduce your problem using the CTX images you were working with (thanks to your attachment I was able to identify and download) as well as other datasets.  Your attachment ultimately gave me a clue when I was having trouble reproducing the problem myself.  It appears the problem has to do with the scale and thus  the zoom factor of the images involved.  The plot tool table reports N/A for some images when all of the linked images do not have the same exact zoom as reported in the title bar for each image window after the file name (it's the @ some number % part).  I noticed in your attachment that not all of the images were the same scale so when I set them to the same scale and plotted another line the table makes more sense.

You must be careful when you set scale especially if you are using the Find tool to locate and center on a common feature in all of your images.  By default the "Sync Scale" option is selected and all linked images of differing resolution will be re-scaled to match the image you selected the feature on resulting in different zoom factors.  I recommend you use the following work around until we can get this problem fixed:

1) link the images
2) use the find tool to center on a common feature
3) use the zoom tool to scale the images to 1:1 and then zoom in/out as needed
4) use the spatial plot tool

This should work but if it doesn't please let us know.

Lynn