Peter,
> > > BTW, an observation. When reading in this WCS NDF creates a very long
> > > LutMap to describe the axis values (8192 elements) which I pass into and
> > > then back out of various channels, this takes a significant amount of time
> > > (even on my Athlon 2800+), when in actual fact the LutMap can be simplied
> > > to a WinMap. Any chance of NDF GTWCS performing this simplication before
> > > the WCS gets out of NDF?
> >
> > One could argue possibly that NDF_GTWCS should return the FrameSet stored
> > using NDF_PTWCS without any change, and that the decision to simplify or
> > not to simplify should be the responsibility of the calling code. Is there
> > any problem with your code calling AST_SIMPLIFY after calling NDF_GTWCS?
>
> Yes, I'm not interested in the actual mapping, I want the FrameSet. To do
> this job I'd have to decompose the FrameSet and then rebuild it.
You can use AST_SIMPLIFY on FrameSets as well as Mappings (since FrameSet
inherits from Mapping). The FrameSet implementation of AST_SIMPLIFY
simplifies all the Mappings in the FrameSet.
> Anyway by
> the time I see this it's too late for the efficiency gains to useful
> (having passed an array of 8192 double precision values into strings once
> already).
Can you spell out the chain of format conversions which GAIA performs?
David
----------------------------------------------------------------------
Dr David S. Berry ([log in to unmask])
STARLINK project | Centre for Astrophysics
(http://www.starlink.ac.uk/) | University of Central Lancashire
Rutherford Appleton Laboratory | PRESTON
DIDCOT | United Kingdom
United Kingdom | PR1 2HE
OX11 0QX Tel. 01772 893733
01257 273192
|