python迭代調(diào)用內(nèi)置函數(shù)計時比較（上）

python迭代工具自動調(diào)用迭代對象next方法，對迭代對象進(jìn)行遍歷。

python的for循環(huán)、列表解析、map方法、生成器表達(dá)式、生成器方法都是迭代工具。

python可迭代對象包括：字符串、列表、元組、字典、集合、range、enumerate、文件等。

1.1 python迭代調(diào)用內(nèi)置函數(shù)計時比較

描述

timetool.py：計時模塊，循環(huán)調(diào)用func函數(shù)1000次，返回使用時間和最后一次調(diào)用結(jié)果。

time.perf_counter()：返回性能計數(shù)器的值，單位為秒。兩次調(diào)用之間的差值用于計時。

timeiterenv.py：各種迭代環(huán)境調(diào)用內(nèi)置函數(shù)ord()，返回列表。對各種迭代函數(shù)調(diào)用計時模塊的計時函數(shù)進(jìn)行計時，將計時結(jié)果存放在列表，并且按從低到高的順序?qū)τ嫊r結(jié)果進(jìn)行排序。

sorted()：key = lambda x:x[1]，按自定義鍵函數(shù)進(jìn)行排序，x為sorted的第1個參數(shù)對應(yīng)排序?qū)ο螅瑇[1]表示按第1個索引值進(jìn)行排序，本例中對應(yīng)函數(shù)計時結(jié)果。

通過例子得出：迭代環(huán)境調(diào)用內(nèi)置函數(shù)耗時從低到高的順序為：

示例

# timetool.py
import time
reps = 1000
repslist = range(reps)

def timer(func,*pargs,**kargs):
   begin = time.perf_counter()
   for i in repslist:
       ret = func(*pargs,**kargs)
   usetime = time.perf_counter() - begin
   return (usetime,ret)

# timeiterenv.py
import sys,timertool
s = '梯閱線條tyxt'*1000

def forloop():
   res = []
   for x in s:
       res.append(ord(x))
   return res
   
def listComp():
   return [ord(x) for x in s]

def mapCall():
   return list(map(ord,s))
   
def genExpr():
   return list(ord(x) for x in s)
   
def genFunc():
   def gen():
       for x in s:
           yield ord(x)
   return list(gen())
   
print(sys.version)
reslist=[]
for test in (forloop,listComp,mapCall,genExpr,genFunc):
   usetime,result = timertool.timer(test)
   reslist.append((test.__name__,usetime,result[0],result[-1],len(result)))

print('-'*33)        
reslistsort=sorted(reslist,key = lambda x:x[1])
for L in reslistsort:
   print('%-9s:%.5f=>[%s....%s....%s]'%(L[0],L[1],L[2],L[3],L[4]))
print('-'*33)

# 迭代調(diào)用內(nèi)置函數(shù)計時比較結(jié)果
# 3.7.8 (tags/v3.7.8:4b47a5b6ba, Jun 28 2020, 07:55:33) [MSC v.1916 32 bit (Intel)]
# ---------------------------------
# mapCall :0.20925=>[26799....116....8000]
# listComp :0.42197=>[26799....116....8000]
# genFunc :0.57103=>[26799....116....8000]
# genExpr :0.57259=>[26799....116....8000]
# forloop :0.66177=>[26799....116....8000]
# ---------------------------------

1.2 python迭代調(diào)用用戶函數(shù)計時比較

描述

python各種迭代環(huán)境調(diào)用用戶函數(shù)ord(x)+1，進(jìn)行計時比較。

timertool.py不變，修改timeiterevn.py即可。

通過例子得出：迭代環(huán)境調(diào)用用戶函數(shù)耗時從低到高的順序為：

示例

# timeiterevn.py
import sys,timertool
s = '梯閱線條tyxt'*1000

def forloop():
   res = []
   for x in s:
       res.append(ord(x)+1)
   return res
   
def listComp():
   return [ord(x) for x in s]

def mapCall():
   return list(map(lambda x:ord(x)+1,s))
   
def genExpr():
   return list(ord(x)+1 for x in s)
   
def genFunc():
   def gen():
       for x in s:
           yield ord(x)+1
   return list(gen())

commstr = '# '
print(commstr+str(sys.version))
reslist=[]
for test in (forloop,listComp,mapCall,genExpr,genFunc):
   usetime,result = timertool.timer(test)
   reslist.append((test.__name__,usetime,result[0],result[-1],len(result)))

print(commstr+'-'*33)        
reslistsort=sorted(reslist,key = lambda x:x[1])
for L in reslistsort:
   print(commstr+'%-9s:%.5f=>[%s....%s....%s]'%(L[0],L[1],L[2],L[3],L[4]))
print(commstr+'-'*33)
# 調(diào)用用戶函數(shù)計時比較結(jié)果
# 3.7.8 (tags/v3.7.8:4b47a5b6ba, Jun 28 2020, 07:55:33) [MSC v.1916 32 bit (Intel)]
# ---------------------------------
# listComp :0.50272=>[26799....116....8000]
# genExpr :0.83316=>[26800....117....8000]
# genFunc :0.85477=>[26800....117....8000]
# forloop :0.94426=>[26800....117....8000]
# mapCall :0.96591=>[26800....117....8000]
# ---------------------------------