Idea from Gailin: I think it would be good to add an automatic diff of the output data, to flag regions and times with large changes and identify the magnitude and sign of those changes. That way we can tell people exactly what data was impacted, and also use it for our own QA to check if changes are expected or not.
Idea from Gailin: I think it would be good to add an automatic diff of the output data, to flag regions and times with large changes and identify the magnitude and sign of those changes. That way we can tell people exactly what data was impacted, and also use it for our own QA to check if changes are expected or not.