Include sigma av as another parameter for starwave to use.

[joycelin1123]

Add in sig_av as another parameter. In starwave.py, get_cmd and sample_cmd have been updated to include this change. In parameters.py, make_params has been updated with an extra line to include this.

[mgennaro]

It would seem that this PR also includes the export of other quantities (sdict) from the get_cmd method. Please clean that up and submit a separate PR for that, so it's documented.
Also, please check what happens if one sets the sigma_av to 0. That will be a typical use case (i.e. no differential reddening) and it would be nice to be able to fallback to the original behavior of no sigma_av. If sigma_av == 0 causes problems with the normal distribution being degenerate/undefined, then please set the default value to something small (eg 1e-5) which is effectively zero, but first check that such small number indeed does not cause problems with the normal distribution.

[mgennaro]

Looking a little closer into this PR. While it is a good idea to have a "sigma_av" to allow for (a gaussian form of) differential reddening, I would suggest withdrawing it and rearranging things to use a truncated gaussian, in order to avoid the extinction from assuming negative values: https://docs.scipy.org/doc/scipy/reference/generated/scipy.stats.truncnorm.html