dust model test
changhoonhahn opened this issue · comments
At the moment we're using a simple Calzetti dust attenuation. We need to test whether this biases our parameter constraints by comparing to a model with a more flexible dust model. The more flexible dust model has already been implemented under model = fsps_complexdust.
@kgb0255 After you've verified that you can reproduce nb/DESI_speculator_comparison.ipynb. Could you re-run the MCMCs for fsps and fsps_complexdust for the first 10 galaxies?
@changhoonhahn Yes. Just for the record, I have fsps version 0.3.0 and I git pulled the updated repo. I believe these two are the only updates in our pipeline. I queued the chains for both fsps and fsps_complexdust. The old chains are at /global/cscratch1/sd/kgb0255/gqp_mc/mini_mocha/ispeculator/james/before_moving.
@kgb0255 Could you post the figure you had in your DESI research forum talk? The one that showed the inferred properties with simple dust vs complex dust.
Yup!
Simple dust model
Complex dust model
