ray-ease 0.1.0

This package is a convenient Ray wrapper that enables the utilization of Ray decorated functions and actors as if they were regular local functions. With this tool, your program can seamlessly run in both parallel and serial modes without requiring any code modifications. This capability is particularly advantageous during the debugging phase, as parallelizing code may inadvertently introduce unnecessary complexities and overhead.

ray-ease

This package is a convenient Ray wrapper that enables the utilization of Ray decorated functions and actors as if they were regular local functions. With this tool, your program can seamlessly run in both parallel and serial modes without requiring any code modifications. This capability is particularly advantageous during the debugging phase, as parallelizing code may inadvertently introduce unnecessary complexities and overhead.

Installation

$ pip install ray-ease

Usage

Effortlessly parallelize your code by simply decorating your functions or classes with the parallelize decorator. Retrieve the results using the retrieve_parallel_loop function. This enables you to parallelize your code with Ray if it's been explicitly initialized or run it serially without any overhead from Ray.

[!NOTE]
In order to correctly decorate functions and classes when parallelizing, ray.init(...) should declared in your script before defining any of the decorated functions or classes. We recommend you to use a environment variable to initialize Ray or not when needed.

Running a Task

import ray
from ray_ease import parallelize, retrieve_parallel_loop

ray.init()

# Define the square task.
@parallelize
def square(x):
    return x * x

# Launch four parallel square tasks.
futures = [square(i) for i in range(4)]

# Retrieve results.
print(retrieve_parallel_loop(futures))
# -> [0, 1, 4, 9]

See Ray version for comparison.

Calling an Actor

import ray
from ray_ease import parallelize, retrieve_parallel_loop

ray.init()

# Define the Counter actor.
@parallelize
class Counter:
    def __init__(self):
        self.i = 0

    def get(self):
        return self.i

    def incr(self, value):
        self.i += value

# Create a Counter actor.
c = Counter()

# Submit calls to the actor. These calls run asynchronously but in
# submission order on the remote actor process.
for _ in range(10):
    c.incr(1)

# Retrieve final actor state.
print(retrieve_parallel_loop(c.get()))
# -> 10

See Ray version for comparison.

Comparison with and without Ray

Parallel computation with Ray (see base example):

import ray
import time
from ray_ease import parallelize, retrieve_parallel_loop

ray.init(num_cpus=4) # Initialize Ray and specify this system has 4 CPUs.

@parallelize
def do_some_work(x):
    time.sleep(1) # Replace this with work you need to do.
    return x

start = time.time()
results = retrieve_parallel_loop([do_some_work(x) for x in range(4)])
print("duration =", time.time() - start)
print("results =", results)

This yields the following output:

duration = 1.0233514308929443
results =  [0, 1, 2, 3]

As opposed to serial computation, by commenting import ray and ray.init(num_cpus=4) lines:

# import ray
import time
from ray_ease import parallelize, retrieve_parallel_loop

# ray.init(num_cpus=4) # Initialize Ray and specify this system has 4 CPUs.

@parallelize
def do_some_work(x):
    time.sleep(1) # Replace this with work you need to do.
    return x

start = time.time()
results = retrieve_parallel_loop([do_some_work(x) for x in range(4)])
print("duration =", time.time() - start)
print("results =", results)

The outputs provide confirmation that the execution was carried out sequentially, taking approximately four times longer than before:

duration = 4.021065711975098
results =  [0, 1, 2, 3]

(back to top)

Contributing

Interested in contributing? Check out the contributing guidelines. Please note that this project is released with a Code of Conduct. By contributing to this project, you agree to abide by its terms.

(back to top)

License

ray-ease was created by Arthur Elskens. It is licensed under the terms of the MIT license. See LICENSE for more information.

(back to top)

Useful links

Repository: https://github.com/aelskens/ray-ease

(back to top)