SRC workflow example: Spectral line finding and classification

Input data: Spectra (frequency vs amplitude)

Output data: Spectra with marked positions of peaks above a (defined?) threshold. CSV file of frequency, peak, FHWM. 

Software involved: Basic Python libraries. Also could be a CNN workflow.

Workflow steps: Very basic: run a script to find peaks after the user has defined thresholds. Plot peaks on the spectra. Extended: Run a CNN to classify optical spectra.

State of existing workflows: Alex has a very simple script that finds peaks. Would be very easy to reproduce. Distributing the data and creating a workflow that searches would be a moderate amount of work. A more detailed (science-ready) workflow needs to be implemented for this in general - Shari. Alex once had students that developed a CNN that classified millions of optical spectra (sub categories of stars and galaxies)

Architecture and hardware: Multithreaded CPU. Minimal RAM. Data can be distributed.