Parallel emcee-based code to study planetary atmospheres.
Python modules needed: dill, schwimmbad, emcee, json, runpy, numpy, corner, mpi4py, lmfit. Thus far, the models are based on petitRADTRANS (https://petitradtrans.readthedocs.io/en/latest/index.html)
The code creates a posterior class for every walker of the MCMC algorithm, that contains a model, a prior and a likelihood.
Each iteration draws a prior following the prior probability and the Markhov chain iteration, evaluates the model from the prior parameters and calculates its likelihood.
A launch script for titan is provided in launch.sh