[daip] [#4188]: blanking in GBT maps

[Eric Greisen]

Many thanks for the FITS file - it will help me test things.

I ran SDGRD on your data and am shocked at how smart it is.  It sets the 
reference declination to 0 at a correct (way offset) pixel location and 
makes a proper image whether or not one specifies APARM.

But XMOM cannot be run blindly - I tried it after TRANS and found that 
no spatial pixel has a legitimate moment if one allows all image values. 
  All of them find that the predominant velocity channel is either < 1 
or > N (probably all the < 1).  Moment analysis assumes a real and 
positive signal - with a lot of channels at values < 0, it fails.  In 
fact, the times I have done moment analysis I have blanked the cube very 
carefully so that channels that are clearly noise are omitted.  The 
OPTYPE 'MAX' avoids some of these issues and makes useful images for you.

With your imaging parameters 1543 of the 1600 image pixels have values - 
the corners are missing a bit.

thanks

Eric Greisen