What is the place to document "the meaning of these measurements"? Jonathan Sick or John Swinbank?

I’m just jumping in here, missing some of the background... plus the fact that we don’t have this type of documentation well implemented yet. So I think I’m looking for an interim solution that will ensure we can successfully move to the final solution.

Given that, these forced photometry measurements are coming from a specific task? Or is it a measurement plugin? (I haven’t yet studied how the measurement architecture works, this is a to-do for me) What I’d like is to have this documentation appear first in the docstring closest to where the measurement is implemented. You can use the “Notes” section of the numpydoc docstring. This won’t solve the documentation problem for users, but at least it will be written down so that someone like me can sweep up the information when we figure out how to document catalogs and measurement plugins.

Does that sound good? I’m very much open to feedback because this is an area of investigation for me.

Thanks Jonathan Sick, that's helpful. I added the following to the ForcedMeasurementTask class docstring:

    Notes

    -----

    *Forced* measurement means that the plugins are provided with a reference

    source containing centroid and/or shape measurements that they may use

    however they see fit. Some plugins can use these to set the location and

    size of apertures, but others may choose to ignore this information,

    essentially performing an unforced measurement starting at the position

    of the reference source (which may nevertheless be useful for certain

    investigations). Knowing how the plugin uses the reference information is

    essential to interpreting its resulting measurements. Typically, centroid

    and shape measurement plugins (e.g., ``SdssCentroid`` and ``SdssShape``)

    are performing unforced measurements.

Lauren MacArthur, do you think that is sufficient?

Looks good Paul, thanks.  I wouldn’t mind an extra comment as to why/when one may prefer the “unforced” measurements from the “forced” catalog as opposed to just using the “unforced” measurements from the “unforced” catalog (but feel free to just stick with what you’ve got if you think that’s self-evident and/or up to the user to discern).

I think the specific uses are up the individual users.

Thanks, all, for your help.

Merged to master.