[News] Re: [Issues] Chip class modifications

[John P. McFarland]

Hi All,

The new Chip class modifications described below have been applied as I didn't 
receive any comments that prevent their implementation.  For those to who it 
applies, please update your checkouts with 'cvs -q update -dPA' to make sure 
you get all the modifications.

Please report any possible Chip-related bugs to the Issues list (the News
list is NOT the place for them).  Thank you.

Cheers,


-=John


On Tue, 20 Feb 2007, John P. McFarland wrote:

> Hi All,
>
> In order to allow inclusion of instruments with multiple CCD geometry 
> configurations into Astro-WISE, the Chip class will be fundamentally 
> modified. This e-mail is to notify the community of these impending changes, 
> particularly for those who write scripts that access Chip methods directly. 
> If you are not interested in these changes, read no further as all the 
> changes necessary to the main opipe checkout will be made in the near future 
> and will be transparent to most users.  Below is a detailed description of 
> the changes.
>
>
> The NAXIS1 and NAXIS2 attributes of Chip will be moved from the chip 
> attribute of BaseFrame-derived classes (where applicable) to attributes of 
> the Frame itself for _all_ objects in the database.  These values will be 
> taken directly from the associated files.  Where the file does not exist, the 
> values will be -1.  The procedure is equivalent to setting all RawFrame 
> NAXIS1, NAXIS2 values to chip.NAXIS1, chip.NAXIS2, respectively, and to 
> setting all non-RawFrame NAXIS1, NAXIS2 values to the values returned by the 
> chip.get_trimmed_size() method.
>
> The PRSCX, PRSCY, OVSCX, OVSCY values will be moved from the chip attribute 
> of RawFrame-derived classes, to attributes of the Frame itself (again for all 
> objects and from the associated files).  The methods:
>
> get_trim_region
> get_trimmed_size
> get_prescan_x_region
> has_prescan_x_region
> get_prescan_y_region
> has_prescan_y_region
> get_overscan_x_region
> has_overscan_x_region
> get_overscan_y_region
> has_overscan_y_region
>
> will be moved from Chip to RawFrame as most of the attributes they use will 
> be moved there.
>
> It should be noted that one major use of these methods is of get_trimmed_size 
> to determine the actual size of a non-RawFrame.  In most cases, this method 
> will no longer be necessary as the desired values will be in the NAXIS1 and 
> NAXIS2 attributes of the Frame itself.
>
> Furthermore, RegriddedFrame and CoaddedRegriddedFrame will gain the NAXIS1, 
> NAXIS2 attributes, allowing methods requiring image sizes (in pixel- or 
> world-coordinates) to access these values directly.  There will no longer be 
> a need to retrieve a possibly enormous file just to determine its geometry.
>
> All methods of the classes and Tasks in the opipe checkout have been modified 
> locally to reflect the above described changes and have been tested to the 
> fullest extent of the unittests.  They will be committed once the database 
> administrators are satisfied that the system is ready for them.  Stay tuned 
> for the announcement of the implementation of these changes, perhaps as early 
> as this Thursday.
>
> Thank you for your attention.
>
> Cheers,
>
>
> -=John
> _______________________________________________
> Issues mailing list
> Issues at astro-wise.org
> http://listman.astro-wise.org/mailman/listinfo/issues
>
>