README for tictoc Repository

Welcome to the tictoc repository! This repository is dedicated to providing a comprehensive benchmarking tool that leverages the bench_dict object. This tool is designed to facilitate the measurement of execution time for various steps in your code, allowing for a detailed analysis of performance bottlenecks and efficiency improvements.

Features

• Step-wise Time Measurement: Utilize the benchdict object to measure the time taken by individual steps within your code.
• Global Step Tracking: Track the time for overarching global steps, providing a broader view of your code's performance.
• Automatic Data Saving: Automatically saves the mean times of each step to a CSV file, making data analysis and sharing straightforward.
• Visualization: Generates a bar plot for the mean time per step and a time series for each global step, offering visual insights into your code's performance.

Usage

Here's a quick start guide to using the BenchDict tool in your projects:

Importing Required Modules

First, import the necessary modules from the tictoc package, along with other required libraries:

from tictoc import bench_dict
from tictoc.basic import count_down_clock, timed_counter
import time
import random

Using the Countdown Clock

You can use the count_down_clock to create a countdown timer. This example creates a 4-second countdown:

count_down = count_down_clock(4)

# check if the timer is done
while not count_down.completed():
    # print how much time is remaining for the countdown to complete
    print('time left', count_down.time_left())
    time.sleep(1)

Benchmarking with BenchDict

Initialize the timed_counter, this class can be used to obtain counts per second and use the bench_dict object to benchmark global steps and individual steps within those global steps:

t_counter = timed_counter()
# start the timer of our counter
t_counter.start()

# reset the timer and the count of our timed counter
t_counter.reset()
for i in range(4):
    bench_dict['g'].gstep()  # Start of a global step for process 'g'
    bench_dict['g2'].gstep()  # Start of a global step for process 'g2'
    time.sleep(1 + random.random())  # Simulate a task
    bench_dict['g'].step('1')  # Mark the end of step 1 within the global step for process 'g'
    bench_dict['g2'].step('1')  # Mark the end of step 1 within the global step for process 'g2'
    bench_dict['g2'].gstop()  # End of the global step for process 'g2'
    time.sleep(1 + random.random())  # Simulate another task
    bench_dict['g'].step('2')  # Mark the end of step 2 within the global step for process 'g'
    bench_dict['g'].gstop()  # End of the global step for process 'g'
    t_counter.count()
    # print count/time from the timed_counter
    print('frequency', t_counter.get_frequency(), 'hz')
# stop the timer of timed_counter
t_counter.stop()

print('frequency', t_counter.get_frequency(), 'hz')
# save results of the benchmark into a file called performance_*timestamp*
bench_dict.save()  # Save the benchmark results