Code for generating plots running times for PR

running_times_plots.py

from random import randint
from timeit import timeit
import pandas as pd
import numpy as onp
import matplotlib.pyplot as plt


def unzip(tuples):
    return tuple(zip(*tuples))


def unzip2(xys):
    xs = []
    ys = []
    for x, y in xys:
        xs.append(x)
        ys.append(y)
    return tuple(xs), tuple(ys)



def generate_items(n, k):
    # n = number of items, k = size of items
    return [tuple(randint(0, 9) for _ in range(k)) for _ in range(n)]

# simple example, 3 pairs:
generate_items(n=3, k=2)  # --> [(9, 5), (2, 4), (0, 0)]

# generate grid:
results = []
for n in onp.logspace(1, 3, 10, dtype='int32'):
    items = generate_items(n=n, k=2)
    assert unzip2(items) == unzip(items)
    result_unzip2 = %timeit -q -o unzip2(items)
    result_unzip = %timeit -q -o unzip(items)
    results.append({
        'n': n,
        'unzip2': result_unzip2.average,
        'unzip2_lower': result_unzip2.average - 1.96 * result_unzip2.stdev,
        'unzip2_upper': result_unzip2.average + 1.96 * result_unzip2.stdev,
        'unzip': result_unzip.average,
        'unzip_lower': result_unzip.average - 1.96 * result_unzip.stdev,
        'unzip_upper': result_unzip.average + 1.96 * result_unzip.stdev,
    })

results = pd.DataFrame.from_records(results).set_index('n')

# plot results
ax = results.plot(y=['unzip2', 'unzip'], style='-o', logx=True, logy=True, figsize=(10, 5))
ax.fill_between(x=results.index, y1=results.unzip2_lower, y2=results.unzip2_upper, alpha=0.25)
ax.fill_between(x=results.index, y1=results.unzip_lower, y2=results.unzip_upper, alpha=0.25)
ax.set_ylabel('seconds', rotation=0, ha='right')
ax.set_xlabel('number of pairs', rotation=0)
ax.set_title('Running time')

plt.savefig('running_times_plots.svg')
plt.savefig('running_times_plots.png')

Output image: