Compare criterion results in text mode
Collect benchmarks using --csv flag to criterion benchmarks.
Beware, criterion appends the results, so CSV file may be "corrupted" that way.
% cabal run -- criterion-cmp old.csv long-name-new.csv
Benchmark old long-name-new
alter 1.027e-3 0.990e-3 -3.62%
delete 2.070e-4 2.157e-4 +4.19%
foldlWithKey 0.945e-3 0.950e-3 +0.50%
foldlWithKey' 2.391e-5 2.375e-5 -0.69%
foldrWithKey 0.458e-7 0.460e-7 +0.37%
fromAscList 1.820e-4 1.828e-4 +0.45%
fromDistinctAscList 1.026e-4 1.054e-4 +2.74%
fromList 2.501e-4 2.541e-4 +1.61%
insert 2.529e-4 2.557e-4 +1.14%
insertLookupWithKey empty 1.023e-3 1.021e-3 -0.21%
insertLookupWithKey update 2.471e-3 2.481e-3 +0.40%
insertWith empty 2.685e-4 2.753e-4 +2.52%
insertWith update 1.189e-3 1.227e-3 +3.24%
insertWith' empty 2.756e-4 2.798e-4 +1.52%
insertWith' update 1.040e-3 1.069e-3 +2.76%
insertWithKey empty 2.660e-4 2.703e-4 +1.59%
insertWithKey update 1.186e-3 1.222e-3 +3.06%
insertWithKey' empty 2.699e-4 2.745e-4 +1.70%
insertWithKey' update 1.032e-3 1.066e-3 +3.36%
lookup 2.947e-4 3.192e-4 +8.34%
map 0.812e-4 0.810e-4 -0.25%
mapMaybe 1.309e-4 1.381e-4 +5.48%
mapMaybeWithKey 1.309e-4 1.382e-4 +5.55%
mapWithKey 0.906e-4 0.908e-4 +0.17%
minView 0.648e-7 0.562e-7 -13.37%
update 0.916e-3 0.969e-3 +5.73%
updateLookupWithKey 1.566e-3 1.621e-3 +3.48%
Geometric mean 1.858e-4 1.885e-4 +1.47%