After a failed experiment in GitHub organization settings, this is the place to ask to have admin privileges to the repo.

Here is a first attempt to define a suitable Sprint Week project based on the simulation files provided by @esheldon. Please add any suggestions or comments below, or define a separate project in a new issue.

The project is to Study the multiplicative biases of the MOF pipeline due to blending as a function of (1) nearest neighbor distance and (2) nearest neighbor flux.

In more detail, p.14 of @esheldon's Nov. 6 talk reports ~2% multiplicative biases using übserseg only, but does not show how biases depend on the properties of individual blends.

The type of plots we are aiming for could be compared with these ones from Samuroff++ 2017 (see also Simon Samuroff's talk from previous week):

This issue to is record useful into about the simulation data files that @esheldon has provided at nersc in /global/projecta/projectdirs/lsst/groups/WL/users/esheldon/nbrsim-outputs/collated/

So far, we only read one file. We should try to read them all and just repeat (combine) the results from the different files.

A next step is to see how the plots defined in #2 are affected by selection effects. @fjaviersanchez thinks magnitude cuts are the primary selection effect @esheldon wants to investigate, so I'll start with those. Other important ones would be cuts on size and SNR.

We should try to filter out the stars. @esheldon said that since the stars have a pretty high SNR it should be clear from the plot snr vs size where these stars live.