Find the lowest cost sequence of edits between two trees
MIT License
Bot releases are hidden (Show)
Published by brunocodutra almost 3 years ago
#31, #32 Dramatic performance improvements across all scenarios measured, from narrow-deep to wide-shallow trees.
cargo bench -- --baseline v0.2.0
n-ary tree diff/depth=7/breadth=2
time: [3.7973 us 3.8099 us 3.8222 us]
change: [-9.2141% -8.4377% -7.7228%] (p = 0.00 < 0.05)
Performance has improved.
n-ary tree diff/depth=3/breadth=6
time: [29.240 us 29.470 us 29.701 us]
change: [-26.040% -25.022% -23.866%] (p = 0.00 < 0.05)
Performance has improved.
n-ary tree diff/depth=2/breadth=15
time: [46.019 us 46.431 us 46.847 us]
change: [-17.322% -15.939% -14.488%] (p = 0.00 < 0.05)
Performance has improved.
n-ary tree diff/depth=1/breadth=255
time: [36.884 us 37.420 us 37.960 us]
change: [-88.071% -86.673% -85.115%] (p = 0.00 < 0.05)
Performance has improved.
#30 Part of trait Node
was split into trait Tree
for ease of use.
Published by brunocodutra about 3 years ago
#21 Relax Node::Kind
to PartialEq
, i.e. Eq
is no longer required.
#23 Change Node::Children
to Deref<Target = [Self::Child]>
from Borrow<[Self::Child]>
.
Published by brunocodutra over 4 years ago