Bayesian fitting

[s-goldman]

Use a prior, maybe the luminosity functions from the Magellanic Clouds, to constrain the luminosities. Then use observations to generate a posterior distribution. This would be a problem for highly-evolved stars which may need a flag in the user interface.

[s-goldman]

This would solve the problems of fitting both spectra and photometry, the unrealistic weighting of data in the far-IR and optical, unrealistic DUSTY models, and finally yield some version of an errorbar

[s-goldman]

I would need an option in the user interface to choose the luminosity function for either LMC, SMC, (or M31)

[s-goldman]

Functions have been created to calculate the grid weights, and next is priors. The way that the code now works is that it creates a grid of luminosities between 1,000 and 100,000 Lsun with a specified sampling in logarithmic spacing. I need to take the specified grid and duplicate and append it for each of the luminosities.

[s-goldman]

Also add option to optimize with emcee.