Run sample repos through the SQALE method and check the results in the same way we checked the initial run (eg, check for correlations, see that one parameter doesn't skew results too greatly).
See how the projects compare to one another, hopefully there is a lot of variation in our sample, so we can see how this measure works given that variation. We probably need to do some demographic analysis on the sample for things like language, project size/complexity, license, etc..
Do we have to run these one at a time or would it buy us anything to figure out a concurrent way to do this? Eg, use a cloud setup and spin up nodes for each run.
Run sample repos through the SQALE method and check the results in the same way we checked the initial run (eg, check for correlations, see that one parameter doesn't skew results too greatly).
See how the projects compare to one another, hopefully there is a lot of variation in our sample, so we can see how this measure works given that variation. We probably need to do some demographic analysis on the sample for things like language, project size/complexity, license, etc..
Do we have to run these one at a time or would it buy us anything to figure out a concurrent way to do this? Eg, use a cloud setup and spin up nodes for each run.