Add error handling so that H must be provided in r filter if object_type = "comet"

[astronomerritt]

The lsstcomet code we use requires H to be in r. So the user should supply H in r if they want cometary magnitudes using the lsstcomet functionality. If they haven't, the code should fail with an informative error message.

The best place for this is probably PPGetMainFilterAndColourOffsets, as this is called very early before any data is read in and checks the column headers in the physical parameters file.

[astronomerritt]

@drewoldag I'd like your feedback on this - lsstcomet requires H to be in the r-filter. However, I hear you're turning the comet activity functionality into a system like the one we already have for light curves, where users can supply their own models, and they might not care what filter H is in.

The options I can see are:

• Add the error handling into PPGetMainFilterAndColourOffsets, knowing it'll need to be changed later.
• Add the error handling into PPCalculateSimpleCometaryMagnitude, which is quite late
• Wait until the new system for comet activity is implemented, then error-handling in PPGetMainFilterAndColourOffsets can check a config file variable to see if PPCalculateSimpleCometaryMagnitude (and thus lsstcomet) is being called

Happy to do whatever will be easiest for you down the road :)

[mschwamb]

Let's deal with this when the lsstcomet class is pulled out

[mschwamb]

This is handled with the upcoming pull request #508