[FEAT] Make more robust python code for plotting

[aaronjridley]

Is your feature request related to a problem? Please describe. There is a very simple plotting code available right now which is not very well written for python and is not very robust. This code should be upgraded so that it follows better coding standards for python and is a bit more robust and able to handle more types of plots.

Describe the solution you'd like The plotting code should:

• Allow users to call the program with the variable name and not just the number (simplified name, too?)
• Allow users to enter an output filename
• Be broken down into more function programming with smaller, reusable, functions
• Allow more plotting types (e.g., zooming in on areas, plotting of 0D or 1D (like specific lat/lon/alt or lat/lon) vs time
• Allow comparisons of model runs (e.g., compare to directories of simulations or simply two files)
• Allow user-defined extension of variable to be plotted (e.g., user creates formula for Joule Heating or something).
• Allow tracing of satellite through code (4D interpolation with comparisons)
• Somehow handles ensembles once they exist (plots the mean values and standard deviations or something....?)

Describe alternatives you've considered Add functionality slowly over time. This issue never closes, since plotting tools are always evolving and are extremely personal.

Additional context So many examples of plots that we could create. So. Many.

[aburrell]

Suggestion: move python code to a separate repository in the AetherModel project. That way people who want Aether but don't use python don't have extra code lying around. And people who just have files don't need to download the entire model to get the visualization tools.

[aaronjridley]

I would tend to not want to do this, since I would rather get the whole package at once. This is really only one "extra" directory, right? And would people use this for anything else beside the support of Aether? I guess it would be much more compelling if this plotting package supported all sorts of different things, like the Chameleon stuff that CCMC puts out. (This makes me think that by default we should support the output of Chameleon files for CCMC software....?) I guess I am always willing to talk about things, though!

Oh - Ben was just talking to us about the idea of linking repositories, so that when you clone one, you automatically get a certain version of another repository as a sub directory. Maybe we could do this? If it is transparent to most users, I would be happy to support this.

[aburrell]

I am not sure how pip support would work with this in an inclusive repository. I don't know how to link repositories (apart from including something as a dependency), but I like the idea. I can see future instances where people wouldn't need to get the model, but want to create plots. Especially once we have a public database of runs.

Apart from the "extra" directory, there will need to be at least installation and testing support.

[qalshidi]

@aburrell I think it would be wise to move to a new repository for pip support as that is how the python inclined would most-likely interface.