Draft: Add Power Law + Dip + Break Model
Current changes:
- add PDB model (with power law pairing and independent pairing) as models g and h
- take in m1 and m2 as parameters instead of m1 and mass ratio
- change priors and injection draw probability accordingly
- some edits in data_simulation.py to accommodate m1 and m2 sampling
- request specific GPUs on LDAS
- normalize in post-processing for PDB models
TO DO:
-
analysis, collection, plotting: allow switch between m1 and m2 vs m1 and mass ratio, for different parameterizations - apparently Colm is already doing this so I just have to wait for his push and make sure the priors are still good
-
plotting: add options for p(m) for pairing function models
**Note: **This MR has some changes that I cannot test until I some fixes are merged into upstream (specifically, the backend imports). I can wait until those changes are available and test them on my fork, but I wanted to make this MR so that you could see what types of changes are coming :)