matlibplotlib.pyplot demo

foo.py

import decimal
import time
from functools import reduce

import matplotlib.pyplot as plt
import random


# generate scale number
def scale_random_number(scale):
    return [random.uniform(10, 20) for _ in range(scale)]


def decimal_deal(numbers):
    return [decimal.Decimal(number) for number in numbers]


#
def operate_time(func, seq):
    start = time.process_time()
    reduce(func, seq)
    end = time.process_time()
    res = end - start
    return res


def add(a, b):
    return a + b


def add_deci(a, b):
    return decimal.Decimal(a) + decimal.Decimal(b)


def graph_show():
    x = list(range(2, 100))
    numbers_ilst = [scale_random_number(l) for l in x]
    numbers_list_pre_decimal = [decimal_deal(numbers) for numbers in numbers_ilst]
    y1 = [operate_time(func=add, seq=numbers) for numbers in numbers_ilst]
    y2 = [operate_time(func=add_deci, seq=numbers) for numbers in numbers_ilst]
    y3 = [operate_time(func=add, seq=numbers) for numbers in numbers_list_pre_decimal]
    with plt.style.context('Solarize_Light2'):
        l1, = plt.plot(x, y1, label='add')
        l2, = plt.plot(x, y2, label='add_decimal')
        l3, = plt.plot(x, y3, label='add_decimal_pre')
        plt.legend(handles=[l1, l2, l3], labels=['add', 'add_decimal', 'add_decimal_pre'])
        plt.title('add')
        plt.xlabel('x Scale of numbers', fontsize=14)
        plt.ylabel('y Operation time', fontsize=14)
        # fig = plt.gcf()
        plt.show()
        # fig.savefig('result.png', dpi=100)


if __name__ == '__main__':
    graph_show()

matplotlib.pylot demo
案例是‘朴素’测试使用内置类型数字加法和decimal加法的性能差异。
ps：注意这里decimal转换消耗其实很很高的，实际上对转换完类型的数字类型进行加法，斜率并不会高出很多。