tic-toc

Measure and track the wall and CPU time of defined scopes in Python.
It's suitable for long-running applications which should be monitored.

Installation

$ pip install tic-toc

Usage

The following examples demonstrate the most simple usage:

Here you see the implicit variant with the with construct:

from tictoc import Timer

with Timer('NAME') as timer:
    print('Scope: ' + timer.name)
    time.sleep(1)

Here you see the explicit variant without the with construct:

from tictoc import Timer

timer = Timer('NAME')
timer.tic()  # or .start()
print('Scope: ' + timer.name)
time.sleep(1)
timer.toc()  # or .end()

The outputs are similar to each other:

# > NAME ...
# Scope: NAME
# < NAME [WALL: 1.0034s] [CPU: 0.0001s]

You can find more extended examples (e.g. with logging, tqdm or pandas) in the examples directory.

Parameters

class Timer(object):
    def __init__(self, name: str = None,
                 format_start: str = '> {name} ...',
                 format_end: str = '< {name} [WALL: {time_wall:.4f}s] [CPU: {time_cpu:.4f}s]',
                 to: Callable[[Any], None] = lambda msg: print(msg)):

• name: str, optional, default: leading hash with four random digits, eg. #0512
• format_start: str, optional, default: '> {name} ...'
• format_end: str, optional, default: '< {name} [WALL: {time_wall:.4f}s] [CPU: {time_cpu:.4f}s]'
• to: Callable[[Any], None], optional, default: lambda msg: print(msg)

You can change the visual string formats and the output method.

Sources

• Answer by Eli Bendersky to the question 'tic, toc functions analog in Python'
• Python Clocks Explained by Nat Dunn
• Python 3.7 perf_counter() nanoseconds by Victor Stinner
• PEP 564: New Time Functions With Nanosecond Resolution
• PEP 560: Core Support for typing module and Generic Types
• Slow and fast methods for generating random integers in Python by Eli Bendersky

License

The module is Open Source Software released under the MIT license.