SWAPI: peak identification of alpha population

[hafarooki]

Summary

The alpha peak is often incorrectly identified. We have a WIP algorithm that seems to do better.

Requirements

Correctly identify the alpha peak. When the alpha peak is not visible because it's washed out by the protons, that should also be corrected reported.

Related tickets

#78 #76 are both related

[jtniehof]

How significant is this for the first data release--are we being really misleading if this isn't done before then, can we catch it enough to flag it, or is this a "must do it right"?

[hafarooki]

After discussion with Jamie: Currently, the algorithm often greatly underestimates the alpha peak by way too much. We did a quick and dirty improvement prototype so far that seems to fix it quite consistently. However, if we can't get it implemented in the pipeline before the milestone, we should at least have a flag to make it clear when the alpha peak has not been correctly identified.