"undoing" dust prior in LOS fits

[joshspeagle]

Currently, the modules for fitting the cumulative reddening along the line of sight do not interface with the addition of the new dust priors I've added into the per-star portion of the code (via fitting from #17). I need to add these in to properly "unweight" the prior to make sure we don't "double-count" it when doing the line-of-sight (LOS) fits.

[joshspeagle]

There are a couple ways this could work in theory. One possibility would be to just save the reddening-specific weights so that users can "unweight" samples afterwards (i.e. perform an importance-correction). This, however, involves saving an addition N_samps numbers, and so increases the memory burden.

Alternately, I could use something similar to what I've implemented for several of the plotting/binning utilities where I recompute the prior using some assumed inputs. This has some overhead involved but also has the benefit of being flexible to different reweighting schemes and keeping consistent argument styles.

[joshspeagle]

Given that:

• the LOS fit is going to be completely reworked in the new 3-D dust modeling approach I want to build up and
• the current approach (without weights) is appropriate for basic population inference

I'm just going to close adding in any relevant weights option for now.