Extended examples and code testing for ompy

Note on usage: ompy is loaded in as a submode. Remember to initialize recursively if you want to use submodules within the submodule: git submodule update --init --recursive

Rendering jupyter notebooks: If you have problems rendering the notebooks directly on git, try accessing them via nbviewer: https://nbviewer.jupyter.org/github/oslocyclotronlab/ompy_further_examples/tree/master/

unfolding:

Comparison of unfolding with ompy and mama. Contains three datasets:

• 145Nd: Comparison of unfolded data + of folded to raw
• 145Nd_artificial: Created a fake raw spectrum with raw_fake = R * raw and unfold this
• 28Si: Comparison of unfolded data + of folded to raw
• folding_efficiency: Check the effect of the efficiency normalization vs normalizing the response R to 1 for each row; otherwise similar idea to 145Nd_artificial

first generation:

Get primaries from all generation spectra and compare to the tru primaries:

• Simple spectrum, idea from Ann-Cecilie (Larsen2011)
• Other simple mock spectra

estimate uncertainty:

Compare different ways to estimate the uncertainty: Either estimate it directly after decomposition, or normalize each ensemble member separately

check_linearity_unfolding_1gen:

A small check on whether the unfolding and first generation method are linear or not. They turn out not to be linear.

development:

• unfolding_fbu: A first test run to show that we can now (easily) use different unfolding algorithms, like fully bayesian unfolding. The priors and other sampler parameters are probably not chosen very smart in this case -- I just wanted to see whether it works in general. And it does so.