Closed czlee closed 8 years ago
See inline comments. Note that we should also divide norm_score in rat-runner by the number of senders.
Otherwise lgtm.
Todo:
I did the new commits on the plots branch, which means that #29 is no longer a superset of this pull request. I think it can still be done automatically (the history isn't that complicated) but if not I'll do appropriate merges when you merge #29.
LGTM
Split of #26 into semantically distinct histories.
Adds a plot.py script that takes a RemyCC file and generates a plot similar to those in the learnability paper, alongside the original data.