CSV files containing pairs of known peak [px] and arc wavelengths [Å or nm] for testing against different fit configurations.
Ideally for as many real world example instruments as we can. Future work should do this with simulated spectra too.
What parameters do we want to explore varying?
- Atlas dropout (missing arc lines)
- Peak dropout (missing peak detections)
- Varying spectrum end tolerance
- Correct/incorrect atlases
*...