New branch comparative_results checked out from unit_test branch to do this (bc unit_test included some bug fixes). Basic idea:
Return a float representing relative ranks.
Ex: mean chat rates for top result was 15/min. For tenth result it was 2/min. So the tenth result is (2/15)*100 = 13% as "good" as the top result for that dataset. Returned as 0.1333.
New json results example:
[
{'startTime': 105.78, 'endTime': 135.449, 'rel_rank': 1.0},
...
...
...
{'startTime': 234.249, 'endTime': 259.889, 'rel_rank': 0.1333}
]
New branch comparative_results checked out from
unit_testbranch to do this (bcunit_testincluded some bug fixes). Basic idea:Ex: mean chat rates for top result was 15/min. For tenth result it was 2/min. So the tenth result is (2/15)*100 = 13% as "good" as the top result for that dataset. Returned as
0.1333.New json results example: