[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

19990517: McIDAS learning guide on waldo

>From: address@hidden
>Organization: St. Cloud State
>Keywords: 199905171536.JAA09507 McIDAS Learning Guide

Alan,

>I have had a disk failure on my desk pc, so email may not
>work.  Am sending this from hobbes; reply to my usual address.

Got it.

>We are working on setting up the tutorial / learning guide on waldo
>Things have gone ok, until a snag appeared last week.  
>The section 'ADDE - getting started ' went fine until near the end.
>
>I can display images GI4 and GV4 ok using IMGDISP;  also get
>a success copying GI4 to TI.5 just as directed by tutorial.
>The alias listings (AKA) so give the expected results.
>
>When doing IMGDISP in step 23
>
>       IMGDISP TI.5 2 SF=YES ...
>
>The background map plots ok but no image plots.  
>
>The command IMGLIST TI.5 ALL shows it is there, but it 
>will not plot for me.  Again, we can plot these images
>using IMGDISP GI4 .... or IMGDISP GV4.
>
>How about a cure?

The things that came to mind were:

o the IMGCOPY in the tutorial specified that the center point for the
  copy is AHN:

IMGCOPY GI4 TI.5 STATION=AHN

  If the ID was misspecified the center for the image that was copied
  might be out of your display space and so the IMGDISP would be
  showing a region with no data.

o it could have been possible (but not likely) that the image display toggle
  was turned off.  This would have the effect of there being an image
  that was navigatable, but no image displayed.

To really check things out, I did the following:

o I performed the steps in the Learning Guide here on a Sun Solaris SPARC
  (i.e. big-endian) machine; everything worked as expected

o logged onto your system and reran the Learning Guide steps including
  the creation of the TI.5 image.  I then pointed my McIDAS session here
  at Unidata to access your MYDATA dataset (through the remote ADDE
  server facility) and loaded TI.5 here.  I confirmed that the image
  is indeed filled with 0 values.  So, the problem is apparently in the
  IMGCOPY reading data from an image that is in big-endian format
  from a little-endian machine.

To get around the above bug, I changed the DATALOC command on waldo for
the BLIZZARD dataset:

<login>
cd workdata
dataloc.k ADD BLIZZARD zero.unidata.ucar.edu

I then reran the IMGCOPY:

imgcopy.k GI4 TI.5 STATION=AHN

Finally, I loaded the image back to my session here at Unidata:

IMGDISP TI.5 2 SF=YES REFRESH='MAP VH 2'

This displayed properly proving that the problem is in the IMGCOPY
on your machine.  This is a bug that I need to report to SSEC.  In
the mean time, you can keep pointing to the BLIZZARD data set here
at Unidata and continue on with the Learning Guide.

Sorry for the problem...

Tom