benchmark

Benchmark

These benchmarks calculates the time it takes to render a composed component and to unmount it afterwards.

You can run the bechmarks using e.g. npm run bench -- ./benchmark/withHandlers.

Results

These are the results running on a Debian 8 VM with 4096 MB RAM and 4 Cores. The Host has a 2,2 GHz Intel Core i7 and runs VMWare Fusion.

withHandlers()

Composing 10x withHandlers()

Source

recompose x 3,264 ops/sec ±2.63% (77 runs sampled)
recompact x 4,855 ops/sec ±2.08% (75 runs sampled)
reassemble x 5,159 ops/sec ±2.89% (70 runs sampled)
Fastest is reassemble

withState()

Composing 10x withState()

Source

recompose x 3,238 ops/sec ±2.08% (79 runs sampled)
recompact x 5,186 ops/sec ±1.75% (77 runs sampled)
reassemble x 5,101 ops/sec ±3.16% (74 runs sampled)
Fastest is recompact,reassemble

withProps()

Composing 10x withProps()

Source

recompose x 9,488 ops/sec ±1.80% (78 runs sampled)
recompact x 5,346 ops/sec ±1.38% (77 runs sampled)
reassemble x 7,799 ops/sec ±1.60% (77 runs sampled)
Fastest is recompose

withPropsOnChange()

Composing 10x withPropsOnChange()

Source

recompose x 3,393 ops/sec ±3.54% (76 runs sampled)
recompact x 5,257 ops/sec ±1.48% (78 runs sampled)
reassemble x 7,081 ops/sec ±1.33% (79 runs sampled)
Fastest is reassemble

Use Case #1: withKeyboardFocus()

Source

recompose x 7,063 ops/sec ±3.70% (76 runs sampled)
recompact x 5,773 ops/sec ±1.93% (74 runs sampled)
reassemble x 6,412 ops/sec ±2.50% (75 runs sampled)
Fastest is recompose