yumeow-timer 0.2.0

Lightweight timing utilities for performance diagnostics

yumeow-timer

Lightweight timing utilities for optional performance diagnostics.

Installation

pip install yumeow-timer

Usage

Basic Timer

from yumeow_timer import Timer

timer = Timer()
for i in range(1000):
    # do something
    timer.add()

print(timer)  # Timer(time=1.23 s, count=1.00 kiter, pace=1.23 ms/iter, speed=813 iter/s)
print(timer.to_str('pace'))  # 1.23 ms/iter

NamedTimer (Multiple Categories)

from yumeow_timer import NamedTimer

timer = NamedTimer()
for i in range(1000):
    if i % 2 == 0:
        timer.add("even")
    else:
        timer.add("odd")

print(timer)  # 1.23 ms/iter (even=1.23 ms/iter[50%]; odd=1.23 ms/iter[50%])
print(timer.to_str('pace', mode_of_detail='pace', mode_of_percent='by_count'))

ParallelTimer (Shared Time)

from yumeow_timer import ParallelTimer

timer = ParallelTimer()
for i in range(1000):
    timer.add("process", n=1)
    timer.add("render", n=1)

print(timer)  # Shows shared time with separate counts

Humanize Functions

from yumeow_timer import humanize_time, humanize_count, humanize_pace, humanize_speed

print(humanize_time(3661))      # 1.02 h
print(humanize_count(1500))     # 1.50 kiter
print(humanize_pace(0.00123))   # 1.23 ms/iter
print(humanize_speed(813))      # 813 iter/s

API Reference

Timer

Basic timer that tracks time and count.

• add(n=1, by="increment") - Add count and elapsed time
• clear(reset_last_add_time=False) - Reset timer
• to_str(mode="pace") - Get human-readable string
• Properties: time, count, pace, speed

NamedTimer

Extends Timer to track multiple named categories with separate time tracking.

ParallelTimer

Extends Timer to track multiple named categories with shared time.

License

MIT License - See LICENSE file for details.