/perf

Performance testing of various nodejs libraries and native functions

Primary LanguageJavaScript

Performance Benchmarks

Tests

These tests have been ran and reported on Mac OSX v10.12.3 15 inch 26 GHz Intel Core i7

====================== sortBy ======================
Slow  lodash      0.74 ops/sec   6 samples
      lazy        1.04 ops/sec   7 samples
      underscore  165 ops/sec    82 samples
      native      200 ops/sec    82 samples
Fast  rambda      5,605 ops/sec  59 samples


====================== sort ======================
Slow  rambda      2.09 ops/sec   10 samples
      lodash      88.89 ops/sec  56 samples
      native      149 ops/sec    71 samples
      underscore  161 ops/sec    79 samples
Fast  lazy        196 ops/sec    81 samples


====================== forEach ======================
Slow  native      1,684 ops/sec   76 samples
      lazy        9,098 ops/sec   35 samples
      lodash      9,300 ops/sec   73 samples
      rambda      80,668 ops/sec  58 samples
Fast  underscore  81,193 ops/sec  43 samples


====================== omit ======================
Slow  native      40.41 ops/sec  53 samples
      lodash      442 ops/sec    64 samples
      underscore  459 ops/sec    83 samples
      rambda      535 ops/sec    87 samples
Fast  lazy        645 ops/sec    81 samples


====================== map ======================
Slow  native      273 ops/sec     84 samples
      lazy        1,575 ops/sec   81 samples
      underscore  7,374 ops/sec   82 samples
      lodash      7,483 ops/sec   59 samples
Fast  rambda      44,020 ops/sec  66 samples


====================== find ======================
Slow  underscore  14,186 ops/sec  60 samples
      lazy        14,545 ops/sec  91 samples
      native      15,076 ops/sec  91 samples
      rambda      19,086 ops/sec  83 samples
Fast  lodash      19,783 ops/sec  53 samples


====================== filter ======================
Slow  native      381 ops/sec     85 samples
      lazy        2,049 ops/sec   77 samples
      lodash      5,023 ops/sec   66 samples
      underscore  6,545 ops/sec   75 samples
Fast  rambda      13,039 ops/sec  75 samples


====================== reduce ======================
Slow  native      557 ops/sec  85 samples
      lazy        739 ops/sec  88 samples
      lodash      850 ops/sec  88 samples
      underscore  887 ops/sec  87 samples
Fast  rambda      894 ops/sec  81 samples


====================== new Promise ======================
Slow  bluebird       0.31 ops/sec       65 samples
      Q              29,922 ops/sec     66 samples
      native         303,416 ops/sec    74 samples
      rsvp           1,128,774 ops/sec  81 samples
Fast  forbesPromise  1,534,114 ops/sec  79 samples