Multi-dimensional Data Access minimal requirements

[Douglas Tody]

On Tue, 11 Mar 2014, Tom McGlynn wrote:

> Douglas Tody wrote:
>> On Tue, 11 Mar 2014, Robert J. Hanisch wrote:
>> 
>>> On 3/11/14 3:49 PM, "Douglas Tody" <dtody at nrao.edu> wrote:
> ...
>>> But either way,
>>> I still think "circle" is not what one wants for a cutout.  It's ok for
>>> discovery, but not for a cutout.
>> 
>> Exactly.  Circle is usable for discovery but not very useful to specify
>> the ROI for a cutout.  Note that in the discovery use case, there are
>> also issues with the size and coverage of the target image - is the
>> center of the target image in the specified circular region, or only
>> part of the image, is it fully contained, etc. (this was the point of
>> the INTERSECT parameter).
>> 
>
> I don't think I agree.  What's wrong with the following?
>
> Hypothetical use case:  I've got 1000 distant galaxies in the Hubble deep 
> field and I want to extract ACS data on each of them.  I'm planning on doing 
> some kind of photometry that needs to go out 5" from the center.  So I 
> request 1000 cutouts as circles with a radius of 5" and the appropriate 
> centers.  I get back 1000 cutouts that are presumably rectangles (probably 
> squares but maybe a pixel or two off).
>
> Seems pretty reasonable to me.  It would be much harder if I have to first 
> calculate the bounding box in RA/Dec for the circle that I need, and it may 
> also be less efficient, since the ACS images may not be nicely aligned with 
> the coordinate system in which I specify my box. So I'd get back larger boxes 
> than I actually needed -- a factor or two more area in the worst case I 
> think.

What you describe is why RANGE is awkward to use for cutouts.

A "circle" specified as POS,SIZE with a single size value does exactly
what you want, but allows for a different size in RA and DEC (or long/lat),
with little if any increase in complexity.

> Recognize that the SIA service is going to have to calculate bounding boxes 
> for a misaligned box even if we are going to only allow boxes. The circle 
> case will almost always be easier for the SIA service to calculate.

Yes, in practice for cutouts one is computing the bounding box (not
circle) around the ROI, keeping whatever orientation the actual image
has.

 	- Doug