Skip to main content
Benchmark snapshot comparing numpy-ts 1.2.0 against native Python NumPy. Run your own via npm run bench.

Benchmark Summary

  • Average speedup: 0.50x vs NumPy
  • Best case: 62.55x
  • Worst case: 0.07x
  • Total benchmarks: 2419
  • Machine: Apple M4 Max (16 cores, 128 GB, arm64)
  • numpy-ts version: 1.2.0

Performance by Category

CategoryAvg SpeedupCountFasterSlower
creation0.92x21590125
arithmetic0.57x29568227
math0.46x14621125
trig0.43x21632184
gradient1.97x22166
linalg1.12x27517798
reductions0.65x416133283
manipulation1.22x231107124
io1.81x664620
indexing0.70x1125557
bitwise0.39x10010
sorting0.83x753045
logic1.15x1387167
statistics0.90x261412
sets2.22x332013
random0.84x461531
polynomials2.77x27189
fft0.81x662442
utilities11.81x440

Performance by DType

DTypeAvg SpeedupMedian SpeedupCount
float640.73x0.66x295
float320.69x0.63x241
float161.07x0.94x220
int640.70x0.67x198
uint640.69x0.62x187
int320.80x0.69x207
uint320.85x0.78x188
int160.86x0.85x188
uint160.85x0.86x184
int80.99x0.97x188
uint80.98x0.98x186
complex1280.57x0.57x74
complex640.50x0.52x63