Using extra task parameters with orchestration

The basic usage of runOrca.py loops over a list of data IDs that the user supplies. This is sufficient for most tasks that operate on a single dataID, or where the task knows how to find the other data it needs. In the case of image differencing, one sometimes needs to also specify the template image ID. E.g., a normal command line might look like imageDifference.py decam_repo --id visit=197391 ccdnum=10 --templateId visit=197395.

Fortunately, runOrca.py will pass along any extra command line parameters that are included with each data ID line. An example input ID list might thus look like:

visit=197379 ccdnum=1 --templateId visit=197371
visit=197379 ccdnum=2 --templateId visit=197371
visit=197379 ccdnum=3 --templateId visit=197371

This does require that each data ID be called as a separate job, (--ids-per-job 1). This appears to be the default in my experience.

1 Like

This was a summary of a discussion with @ktl and @srp. There’s no question to be answered, so I will mark this as a solution.