ultrachronic 0.1.5

Run parallel jobs and save results to json.gz files

Run parallel jobs and save results to json.gz files.

Install

Stable:

pip install ultrachronic

Latest:

pip install git+https://github.com/yoavram/ultrachronic.git

Supports Python 3.4 and 3.5.

Test

Requirements:

pip install nose click

Run:

nosetests tests

Example

Code in do_something.py:

from ultrachronic import jsonify_result, repeat

import click

@jsonify_result
def do_something(arg1, arg2):
    a = 1
    b = 2
    # must return a dict!
    return dict(a=a, b=b)

@click.command()
@click.option('--arg1', default=1, type=int, help='Argument 1')
@click.option('--arg2', default='a', type=str, help='Argument 2')
@click.option('--reps', default=1, type=click.IntRange(1, None), help='Number of repetitions')
@click.option('--cpus', default=1, type=int, help='Number of CPUs to use (<1 for all available)')
def main(arg1, arg2, reps, cpus):
    repeat(do_something, reps, cpus, arg1=arg1, arg2=arg2)

if __name__ == '__main__':
    main()

Usage:

python do_something.py
python do_something.py --reps 10 --cpus 2
python do_something.py --arg1 5 --reps 10 --cpus 1
python do_something.py --arg2 hi --reps 10 --cpus 0

Authors

• Yoav Ram (@yoavram)