It's becoming clear that it is important to track all data changes on the processing object. We can compare raw and processed data to see all changes, but a more granular description of all changes will be really good for the stats we can reconstruct out of that after the fact.
I can possibly see a reason for some sort of data structure that saves the following:
[x] Type: Gap removal, ramp correction, spike removal, whatever.
[ ] Range: Begin and end date? Surely this should be done in blocks.
[ ] Colour for plotting purposes?
[x] Some tags, maybe a category system? Destructive, transformation, substitution, etc?
[ ] Note: Dependent on the change being made. For a substitution maybe the source file/site? Shift amounts, ramp equations, other notes, reasons for change, etc.
It's becoming clear that it is important to track all data changes on the processing object. We can compare raw and processed data to see all changes, but a more granular description of all changes will be really good for the stats we can reconstruct out of that after the fact.
I can possibly see a reason for some sort of data structure that saves the following: