karllark / dirtygrid Goto Github PK

@jchastenet: is this due to not having the interpolation code to expand the sparse spectroscopic grid to the "uniform" photometry grid?

the models results give the full spectrum
need to have a way to get this full spectrum via the dirtygrid class
will have to decide how to access/store this larger volume of results

• Add colors (for now, the scatter plot in PhotDG is all same color if multiple SEDs)
• Update the axis depending on the unit chosen
• Add automatic legend
• Add option to specify an subplot

Comparing the DG results to photometry at different redshifts is clearly desired. Need code to make this straightforward based on on the spectroscopic DG.

The DirtyGrid spectroscopic results are used to create the photometric grid. The code to generate photometry in a list of bands with provided response curves is needed.
@jchastenet: is this something you have and can provide (even if it is bare bones)?

revise original plot_and_print_dirtygrid to be a class
initialization to read in the dirtygrid (photometry only)
functions to:
plot a SED
print a SED
return a SED
return a set of SEDs

need to be able to generate new photometry from an input filter response function

Need to add a component to ask for a conversion from the DGrid initial units (erg/s/micron)

We should include the response curves used to generate the photometric DG. This would allow for the grid to be (re)generated as needed (e.g., different redshift, new DG runs, etc.).

The full spectra are large and not well organized. And located only on central storage in many subdirectories. Jeremy and I have been talking about the need to get them more organized and posted publically. I will work on how to do this in manageable chunks. I'm currently thinking of having separate files for the 'discrete' variables (e.g., dust grain type, star formation type) with all the continuous variables for that set of discrete variables in one file (e.g., tau, sfr). Does this sound like a reasonable organization?

Need to start writing test code.
This will allow for automated travis ci testing.
Complicated as the code is directly coupled to working on the DG files. Maybe make a "small" DG grid and use the astropy remote_data methods.

functions such as photoGet only support scalar inputs, which can decrease program efficiency significantly. As a result SEDs can only be retrieved once a time. Need to add in array input support.