Closed fschr closed 7 years ago
What I intend to do next, is create a branch (one for each e2e
test) where a script is created that outputs a slightly different distribution. This the branch can be compared to master
and we can expect certain p-values, etc. based on how different the distributions are. This allows us to perform end-to-end tests on regress build ...
and regress stat ...
.
All 19 commits are essentially the same. We need 19 commits so we can run
regress --sample-size=19
(that command will checkout the latest 19 commits onmaster
).