philiprehberger-timerfunc
v0.2.0732 downloadsMIT
Measure execution time of any function
$
pip install philiprehberger-timerfuncREADME
philiprehberger-timerfunc
Measure execution time of any function.
Installation
pip install philiprehberger-timerfunc
Usage
Context Manager
from philiprehberger_timerfunc import timer
with timer() as t:
do_work()
print(f"Took {t.elapsed_ms:.1f}ms")
Decorator
from philiprehberger_timerfunc import timed
@timed
def process_data():
... # logs: "process_data took 42.3ms"
@timed(threshold_ms=100)
def api_call():
... # only logs if slower than 100ms
Benchmark
from philiprehberger_timerfunc import benchmark
result = benchmark(my_function, args=(data,), iterations=1000)
print(f"Mean: {result.mean_ms:.2f}ms, P95: {result.p95_ms:.2f}ms")
print(f"Fastest: {result.fastest:.2f}ms, Slowest: {result.slowest:.2f}ms")
print(result) # full stats summary
Comparing benchmarks
from philiprehberger_timerfunc import benchmark, compare_benchmarks
old = benchmark(old_impl, iterations=1000)
new = benchmark(new_impl, iterations=1000)
ratio = compare_benchmarks(new, old)
# ratio > 1 means `new` is faster than `old` (old took proportionally longer)
print(f"new is {ratio:.2f}x faster than old")
API
| Function / Class | Description |
|---|---|
timer() | Context manager returning TimerResult |
@timed / @timed(threshold_ms=0) | Decorator logging execution time |
benchmark(fn, args, kwargs, iterations, warmup) | Returns BenchmarkResult |
BenchmarkResult.fastest | Duration of the fastest run (ms); alias of min_ms |
BenchmarkResult.slowest | Duration of the slowest run (ms); alias of max_ms |
compare_benchmarks(a, b) | Returns b.mean / a.mean; values > 1 mean a is faster |
Development
pip install -e .
python -m pytest tests/ -v
Support
If you find this project useful: