python性能对比

python性能对比之items

#1

#-*- coding:utf8-*-

import datetime

road_nodes = {}
for i in range(5000000):
    road_nodes[i] = {'id':i}

beg_time = datetime.datetime.now()
for key, val in road_nodes.items():
    pass
end_time = datetime.datetime.now()

print "time_scan:", end_time - beg_time

耗时

$python com_1.py
time_scan: 0:00:17.126098

#2

#-*- coding:utf8-*-

import datetime

road_nodes = []
for i in range(5000000):
    road_nodes.append(i)

beg_time = datetime.datetime.now()
for key in road_nodes:
    pass
end_time = datetime.datetime.now()

print "time_scan:", end_time - beg_time

耗时

$python com_2.py
time_scan: 0:00:00.325304

#3

#-*- coding:utf8-*-

import datetime

road_nodes = {}
for i in range(5000000):
    road_nodes[i] = {'id':i}

beg_time = datetime.datetime.now()
for key in road_nodes:
    pass
end_time = datetime.datetime.now()

print "time_scan:", end_time - beg_time

耗时

$python  com_3.py
time_scan: 0:00:00.359197

总结

只对于字典key的遍历，用for key in dic方式　　

python性能对比之in list/dict

#1

from datetime import datetime
a = []
for i in range(50000):
    a.append(i)

aim_list = []
for i in range(1000000):
    aim_list.append(i)

beg_time = datetime.now()
for key in a:
    if key in aim_list:
        pass
end_time = datetime.now()
print "scan:", end_time - beg_time

运行

$python draw.py
scan: 0:00:38.624985

#2　　

from datetime import datetime
a = []
for i in range(50000):
    a.append(i)

aim_list = {}
for i in range(1000000):
    aim_list[i] = None

beg_time = datetime.now()
for key in a:
    if key in aim_list:
        pass
end_time = datetime.now()
print "scan:", end_time - beg_time

运行

$python  draw2.py
scan: 0:00:00.014031

分析

巨大的差异由 aim_list类型导致，一个是list，一个是dict

for i in ***时，list是一个个遍历，效率是O(n);dict是直接找到，效率是O(1)。