Skip to content

Latest commit

 

History

History
30 lines (16 loc) · 2.15 KB

README.md

File metadata and controls

30 lines (16 loc) · 2.15 KB

Original code comes from https://arxiv.org/abs/2112.03864

This is the cleaned, upgraded version. Less WIP.

dovekie.py collects the relevant data and calculates the colour/magnitude slope for each band. It will also run a NUTS mcmc process to determine the best offsets, if asked.

python dovekie.py will not run the mcmc. Instead, it will generate a set of colour/magnitude plots for each filter combination and apply the initial magnitude offsets calculated from the PS1 aperture magntiudes. When finished, it will generate a file called preoffsetsaper.dat, which contains a lot of very helpful diagnostic information, such as the number of stars in each survey, the synthetic colour/magnitude slopes, the data colour/magnitude slopes, and the significance of the diference between them.

mcmc.py has some additional options: DEBUG - a debug key that prints additional information if needs be.

IRSA - This will run an IRSA dust query on each of the surveys that you have specified in surveys_for_chisq. This should only be run after you have generated new observed data to use, at is a bit time consuming, and is disabled by default.

FAKES - if you want to test the validity of the mcmc with fake stars, this will switch mcmc.py to look in an alternative directory, output_fake_apermags.

MCMC - runs the MCMC process, and outputs the chains every 100 steps. This takes time, so best to do in a screen or something.

A lot of information that dovekie.py reads in is stored in DOVEKIE_DEFS.yaml. This file contains input information (which surveys to fit) and output information; e.g. where to store the mcmc chains.

REGENERATE SYNTHETIC MAGNITUDES:

  1. run python loopsyntheticmags_commandlineaper.py without any commands. This will give a list of available surveys to process, with an associated integer.
  2. choose the survey integer and run python loopsyntheticmags_commandlineaper.py INT where INT is your desired number.
  3. If you desire a shift be applied to the filters, add the --SHIFT "np.arange(minval,maxval,binsize)" key in the command line, which will apply shifts based on your input.