[lsst-dm-stack-users] Generating registry.sqlite3

Simon Krughoff simon.krughoff at gmail.com
Wed Jul 24 13:03:19 PDT 2013


Hi James,

I wonder if the files you are processing are not in the correct
directory structure.  From the directory where you are running
genInputRegistry.py, can you send me the output of something like:
$> find . |grep -i fpc |head -n 20
so I can see what structure your files are in?

Thanks,
Simon

On Wed, Jul 24, 2013 at 12:29 PM, James McElveen <jmcelve2 at gmail.com> wrote:
> Hi Simon,
>
>> You only need a registry.sqlite3 file for the data directory.  The
>> file you are downloading is the astrometry_net_data, as you noted, so
>> doesn't need a registry file.
>
> This would have been clear had I properly read the explanation on the Trac site. My apologies! Everything is clear now as far as what the registry file does.
>
>> I'm curious what data you are trying to run and what error you are getting.
>>
>> In general there is a genInputRegistry.py file in the obs package for
>> each camera.  For example, if you have set up obs_sdss (in this case):
>> 1. Go to the directory where the raw data is.  This is where the run
>> directories are (e.g. 4868/).
>> 2. Run:
>> $> $OBS_SDSS_DIR/bin/genInputRegistroy.py .
>
> When I run this, it generates a registry file, but it also gives the following in the standard output:
>
> . ... started
> . ... 0 processed, 0 skipped, 0 unrecognized
> Cleaning up...
>
> So it doesn't look like it's processing anything. Is this expected?
>
> I'm trying to preprocess an SDSS fpC file, but it doesn't seem to be reading in the specific data IDs:
>
>  WARNING: No data found for dataId=OrderedDict([('run', 1033), ('camcol', 2), ('field', 111), ('filter', 'g')])
>
> My guess is that the registry file isn't being generated properly, so the pipeline doesn't know how to process the IDs.
>
> James



-- 
K. Simon Krughoff
University of Washington
Department of Astronomy

Physics-Astronomy Bldg.
3910 15th Ave NE
Room C319
Seattle WA 98195-0002
USA

(206)543-9487



More information about the dm-users mailing list