Currently the parsedata.writecsv function return a dict with (1) the delayed job start time, (2) the corresponding avg carbon intensity over the job runtime and (3) the total carbon intensity over the job. Only (1) is currently used.
The carbon footprint estimation module (carbonFootprint.py) requires (2). To estimate carbon savings, it also needs the average carbon intensity corresponding to running the job right now without waiting. AFAIK - the total carbon intensity (3) is not required.
This PR basically makes 2 changes:
make time_series_conversion.get_lowest_carbon_intensity() return a tuple containing (1) and (2) only - and associated changes in calllers
add a new function parsedata.avg_carbon_intensity() that returns the avg carbon intensity given a job start time and a job runtime. it's used with start = now().
If you're time-crunched, I would say parsedata.avg_carbon_intensity() needs the most reviewing
closes #28
Currently the
parsedata.writecsvfunction return a dict with (1) the delayed job start time, (2) the corresponding avg carbon intensity over the job runtime and (3) the total carbon intensity over the job. Only (1) is currently used.The carbon footprint estimation module (carbonFootprint.py) requires (2). To estimate carbon savings, it also needs the average carbon intensity corresponding to running the job right now without waiting. AFAIK - the total carbon intensity (3) is not required.
This PR basically makes 2 changes:
time_series_conversion.get_lowest_carbon_intensity()return a tuple containing (1) and (2) only - and associated changes in calllersparsedata.avg_carbon_intensity()that returns the avg carbon intensity given a job start time and a job runtime. it's used withstart = now().If you're time-crunched, I would say
parsedata.avg_carbon_intensity()needs the most reviewing