memoizetest.py

from __future__ import print_function

import numpy as np
from datetime import datetime   # timing
from rx import Observable
import random
from concurrent.futures import ProcessPoolExecutor
from memoizepy import memoize
import IPython


@memoize
def longrun_eig(mtrx):
    """ takes a matrix, returns eigenvalue sum """
    eigs = np.linalg.eig(mtrx)[0]
    return sum(eigs)

@memoize
def primes_between(a, b):
    primes = []
    for num in range(a, b):
        if all(num % i != 0 for i in range(2, num)):
            primes.append(num)
    return(len(primes))

def testspeed(num_processes = 6, eigruns = 6):
    adim = 300
    num_processes = 6
    arrays = [np.random.rand(adim * adim).reshape(adim, adim) for _ in range(eigruns)]
    print("longrun_eig not memoized:")
    now_time = datetime.utcnow()
    with ProcessPoolExecutor(num_processes) as executor:
        Observable.from_(arrays) \
            .flat_map(lambda s: executor.submit(longrun_eig, s)) \
            .subscribe(print)
    print("time taken {}".format(datetime.utcnow() - now_time))
    print()
    print("longrun_eig memoized:")
    now_time = datetime.utcnow()
    Observable.from_(arrays) \
        .map(longrun_eig) \
        .subscribe(print)
    print("time taken {}".format(datetime.utcnow() - now_time))
    print()
    test_ranges = [(random.randint(1, 5000), random.randint(5001, 15000)) for _ in range(eigruns)]
    print("primes_between not memoized:")
    now_time = datetime.utcnow()
    with ProcessPoolExecutor(num_processes) as executor:
        Observable.from_(test_ranges) \
            .flat_map(lambda x: executor.submit(primes_between, *x)) \
            .subscribe(print)
    print("time taken {}".format(datetime.utcnow() - now_time))
    print()
    print("primes_between memoized:")
    now_time = datetime.utcnow()
    Observable.from_(test_ranges) \
        .map(lambda x: primes_between(*x)) \
        .subscribe(print)
    print("time taken {}".format(datetime.utcnow() - now_time))

if __name__ == "__main__":
    testspeed()