feat: allow operation by app as well as python class

[john-halloran]

Problem

Some users may prefer an app-like workflow instead of importing a python class, and by default scikit-package includes a pathway for one. But currently, the only function of the "app" is to return the version and a basic help message.

Proposed solution

Since the program is very simple, the app can work much like the class does. Instead of importing the stretched-nmf class, a user might run something like stretched-nmf data_file.xy rho=1e12 eta=610 show_plots=True, which would create output files in the same directory.

[sbillinge]

let's do that. Do you want to do that for this release? It is more work (not too much) but the benefit is that when we roll it out it will be easier to use and may get adopted more easily.

btw, don't we run it on a directory full of files? What kind of data-file.xy (dash not underscore per group standards) contains a full stack of curves? Once we start making an app we can make it more and more fancy, but don't get tempted to do that initially. Have it run on a directory of files and crash if any of the files are bad. then it is up to the user to make sure that only good files are in there. Also, have it plot them so the results are in alpha-numeric order. Again, the user has to figure out filenames so this makes sense, though it usually does if their name is something like <compound-name>-<temperature>K.xy, they will plot vs increasing temperature. But bottom line, make the app simple, an mvp, and we can add features when, and only when, people request them.