[cube access] minutes from telecon (2013-07-03)

[Patrick Dowler]

The purpose of a prototype is not to actually deliver the service with 
SUM as the result. The purpose is for data centres to put the code 
together and see if they can actually do it and what the problem areas 
are. Once that is done, changing the algorithm to INTEG or something 
else is entirely trivial in comparison.

So, forget about the usefulness of SUM or AVG. They are just easy to do. 
The same goes for the quibbling about STC-S and what is in or out. As 
was mentioned in earlier responses, this is just the simplest sub-cube 
extraction with no resampling or pixel operations. Once that is 
implemented, we will be in a position to discuss the actual issues with 
providing such a service... then we can figure out exactly where the 
line is, but not before.

Pat

On 07/24/2013 10:53 AM, Arnold Rots wrote:
> Similarly, the AIPS MOMNT task applies a mask derived from a smoothed
> (in 3-D) version of the cube, but usually it takes some fiddling with its
> parameters to get an optimal result.
> SUM c.s. (including unmasked moments) are pretty useless, because you
> are mainly adding noise.
> That's why I am skeptical about the general usefulness (and urgency) of
> adding these kind of services.
>
> Cheers,
>
>    - Arnold
>
> -------------------------------------------------------------------------------------------------------------
> Arnold H. Rots                                          Chandra X-ray
> Science Center
> Smithsonian Astrophysical Observatory                   tel:  +1 617 496
> 7701
> 60 Garden Street, MS 67                                      fax:  +1
> 617 495 7356
> Cambridge, MA 02138 arots at cfa.harvard.edu <mailto:arots at cfa.harvard.edu>
> USA http://hea-www.harvard.edu/~arots/
> --------------------------------------------------------------------------------------------------------------
>
>
>
> On Wed, Jul 24, 2013 at 11:26 AM, Tim Jenness <tjenness at cornell.edu
> <mailto:tjenness at cornell.edu>> wrote:
>
>
>     On Jul 23, 2013, at 15:47 , Douglas Tody <dtody at nrao.edu
>     <mailto:dtody at nrao.edu>> wrote:
>
>      > Interesting discussion.  My view, having read through all the
>     postings,
>      > follows.
>      >
>      > *** SUM, AVG, INTEG, moments, etc.
>      >
>      > This is an important issue (especially for the more general
>     AccessData
>      > problem, not just the cutout prototype).
>      >
>      >
>
>     …
>
>      >
>      > For the simple "cutout" prototype this is too ambitious, it should be
>      > left to the AccessData prototyping, which we need for our full up
>     cube
>      > access prototyping.  The most we might want to do is define more
>      > precisely what is permitted for something like SUM, e.g., if the
>      > channels are not equal in size, compute the integral instead.
>      >
>
>     I'm not sure I understand. So if you have equal-sized channels you
>     get the SUM but are lacking the information to be able to convert it
>     to an integral. If you have unequal channels you do get the
>     integral. Wouldn't it be easier to give you the integral every time?
>     It seems that the difference between SUM and INTEG is so small that
>     adding explicit support for INTEG is not much of an issue (but make
>     sure the units on the output image are correct). The other moment
>     maps are almost as easy if you use the naive formula.
>
>     More importantly, I find that naive SUM/INTEG are not as useful as
>     you might think because you still can't see that weak line in the
>     sum if there is a lot of baseline adding in lots of 0+/-noise to the
>     SUM. The JCMT pipeline runs a clump-finding algorithm first and
>     masks the baseline regions before calculating all the moment maps.
>     This gives very nice results but I imagine is not something you want
>     to be doing in a general service.
>
>     --
>     Tim Jenness
>
>

-- 

Patrick Dowler
Canadian Astronomy Data Centre
National Research Council Canada
5071 West Saanich Road
Victoria, BC V9A 2L9

250-363-0044 (office) 250-363-0045 (fax)