list dict 性能測試

from random import randint


def load_list_data(total_nums, target_nums):
    """
    從文件中讀取數據，以list的方式返回
    :param total_nums: 讀取的數量
    :param target_nums: 須要查詢的數據的數量
    """
    all_data = []
    target_data = []
    file_name = "fbobject_idnew.txt"
    with open(file_name, encoding="utf8", mode="r") as f_open:
        for count, line in enumerate(f_open):
            if count < total_nums:
                all_data.append(line)
            else:
                break

    for x in range(target_nums):
        random_index = randint(0, total_nums)
        if all_data[random_index] not in target_data:
            target_data.append(all_data[random_index])
            if len(target_data) == target_nums:
                break

    return all_data, target_data

def load_dict_data(total_nums, target_nums):
    """
    從文件中讀取數據，以dict的方式返回
    :param total_nums: 讀取的數量
    :param target_nums: 須要查詢的數據的數量
    """
    all_data = {}
    target_data = []
    file_name = "fbobject_idnew.txt"
    with open(file_name, encoding="utf8", mode="r") as f_open:
        for count, line in enumerate(f_open):
            if count < total_nums:
                all_data[line] = 0
            else:
                break
    all_data_list = list(all_data)
    for x in range(target_nums):
        random_index = randint(0, total_nums-1)
        if all_data_list[random_index] not in target_data:
            target_data.append(all_data_list[random_index])
            if len(target_data) == target_nums:
                break

    return all_data, target_data


def find_test(all_data, target_data):
    #測試運行時間
    test_times = 100
    total_times = 0
    import time
    for i in range(test_times):
        find = 0
        start_time = time.time()
        for data in target_data:
            if data in all_data:
                find += 1
        last_time = time.time() - start_time
        total_times += last_time
    return total_times/test_times


if __name__ == "__main__":
    # all_data, target_data = load_list_data(10000, 1000)
    # all_data, target_data = load_list_data(100000, 1000)
    # all_data, target_data = load_list_data(1000000, 1000)


    # all_data, target_data = load_dict_data(10000, 1000)
    # all_data, target_data = load_dict_data(100000, 1000)
    # all_data, target_data = load_dict_data(1000000, 1000)
    all_data, target_data = load_dict_data(2000000, 1000)
    last_time = find_test(all_data, target_data)

    #dict查找的性能遠遠大於list
    #在list中隨着list數據的增大 查找時間會增大
    #在dict中查找元素不會隨着dict的增大而增大
    print(last_time)

不可變對象 都是可hash的， str， fronzenset， tuple，本身實現的類 __hash__
dict的內存花銷大，可是查詢速度快， 自定義的對象 或者python內部的對象都是用dict包裝的

查找效率：set>dict>list

單次查詢中：看來list 就是O(n)的；而set作了去重，本質應該一顆紅黑樹（猜想，STL就是紅黑樹），複雜度O(logn)；dict相似對key進行了hash,而後再對hash生成一個紅黑樹進行查找，其查找複雜實際上是O(logn),並非所謂的O(1)。O(1)只是理想的實現，實際上不少hash的實現是進行了離散化的。dict比set多了一步hash的過程，so 它比set慢，不過差異不大。