Moose: Molecular Optical Emission Simulation for python

Welcome to Moose, a python package for simulating optical emission spectra for diatomic molecules.

Moose was born out of a need to fit rotational and vibrational temperatures for a large set of data.

For this it uses line-by-line databases, assuming a Boltzmann distribution for rotational and/or vibrational temperatures.

On top of that Moose is intended to be minimal: it provides you some basic tools to do just that, simulate some spectra.

It is up to you to read and sanitize experimental data, that you would like to fit based on these simulations, for instance.

In addition, Moose is quite adaptable: you can use it as a foundation for more elaborate models and analysis.

Put differently, Moose is aimed at helping you: you can integrate and adapt it to your workflow, rather than the reverse.

Additionally, it has only a few, common dependencies (numpy, scipy, pandas) by default.

To get a better grasp of how it works under the hood, see the code reference

MassiveOES

The code that turned into Moose, was intended as a wrapper for MassiveOES, to parallelize it with Dask.

However, the basic desired functionality was easier to implement separately in a stateless manner, rather than probably reworking the entire library.

The goal is not to provide the same functionality as MassiveOES, but instead provide a basis to build your own code on top of.

Moose still uses the same format as MassiveOES for the line-by-line databases, since it makes them interoperable.

You may be better served using MassiveOES or other programs, because of the following:

• Moose does not give you tools to read, correct, or plot spectra
  • There are better, more feature complete libraries for that
• Moose lacks out-of-the-box fitting of multiple excited species not in equilibrium
  • See the examples pages to see how you could use Moose for fitting
  • Or simply build upon Moose for your own custom fitting routines
• The provided fitting examples simply recalculate the simulated spectrum for each fit iteration, which sacrifices some speed.
  • The benefit is that it can be run in parallel with ease.
  • If you don't have a huge set of very similar data or complex fits, this should not be an issue