[daip] subim limitation

[Lawrence Rudnick]

Your question is tricky, and suggests that perhaps we'd be better off 
not doing anything?

   Here's the deal -- when y and z are both spatial coordinates, then 
i'd want to make the same choice of opc, (which i always set to ave), so 
i'd have xinc of 1 (freq axis) and yinc and zinc of n (and have them do 
average)  however, suppose x and y are spatial and z is frequency.  I 
might want to make a different choice.  in that case, I would be better 
off doing subim in two steps.

xinc*yinc*zinc, full average, makes sense to me as a default, and if i 
didn't want to do that, i'd have to do partial subims.  treating zinc 
differently than xinc and yinc would almost certainly create confusion.

given this, i'm almost tempted to say just leave it as is. perhaps try 
out the question on others?

lr


On 11/25/2013 09:48 AM, Eric Greisen wrote:
> Lawrence Rudnick wrote:
>>   I have an enormous cube from FARS, 400x8000x8000.    i want to do a
>> subim with yinc=5,zinc=5, but there's no way to input zinc.
>> transposing and then subim'ing and then transposing back is an option,
>> but the transpose of the initial cube is a real bear.  i've managed a
>> workaround, doing partial subim first, then transpose to get other
>> axis.  since subim works on up to 7d images, incs might be better to
>> also go 7d.
>
> We essentially never have more than 3 so ZINC should be enough.
>
> The real issue is whether you would like the AVE, MIN, and MAX
> operations to work only in every ZINC'th plane or should they function
> in the full XINC x YINC x ZINC cube?  The former is easy, the latter
> will require a re-write of this ancient task but one I would not be
> entirely loathe to do.
>
> Eric

-- 

Lawrence Rudnick
Distinguished Teaching Professor
Minnesota Institute for Astrophysics
School of Physics and Astronomy
University of Minnesota
http://umn.edu/~larry   larry at umn.edu
+1-612-624-3396