云机和主机上运行golang，julia，octave，python，c，java，计算十万，百万，一千万次不同浮点数的正弦值，统计计算耗时

云机和主机上运行golang，julia，octave，python，c，java，计算十万，百万，一千万次不同浮点数的正弦值，统计计算耗时

简介：

实验环境

c语言

Python

Java

golang

julia

实验环境：

本机是win10系统，云机是centos7系统

c语言：

在本机上我使用的是notepad++编辑器，编译是在win10窗口下运行的，采用的是tcc编译的。

c代码：

十万次不同浮点数的正弦值代码

#include<stdio.h>
#include<math.h>
#include<stdlib.h>
#include<time.h> 
#define PI 3.1415
void main(){
	int begintime,endtime; 
    int i = 0;
    float n;
	begintime=clock();  //开始计时
	for(i=0;i<100000;i++)
	{
	n=(rand()%100000)*0.00001;
    printf("%.8f %.8f %.8f\n",n ,sin(n*PI/180),cos(n*PI/180));
}
    endtime = clock(); //结束计时
	printf("\n\nRunning Time:%dms\n",endtime-begintime);
}

本机运行结果：

0.05801000 0.00101244 0.99999949
0.04280000 0.00074698 0.99999972
0.05253000 0.00091679 0.99999958
0.01629000 0.00028431 0.99999996


Running Time:31168ms

运行时间是31168ms

云机上运行结果：

这个在云机上运行要使用 gcc sin.c -o sin -lm，命令不然编译不出来

0.48302999 0.00843012 0.99996447
0.58275002 0.01017043 0.99994828


Running Time:250000ms
(base) [root@VM_0_17_centos 汇报文档1]#

云运行时间250000ms

一百万次不同浮点数的正弦值代码:

#include<stdio.h>
#include<math.h>
#include<stdlib.h>
#include<time.h> 
#define PI 3.1415
void main(){
	int begintime,endtime; 
    int i = 0;
    float n;
	begintime=clock();  //开始计时
	for(i=0;i<1000000;i++)
	{
	n=(rand()%100000)*0.00001;
    printf("%.8f %.8f %.8f\n",n ,sin(n*PI/180),cos(n*PI/180));
}
    endtime = clock(); //结束计时
	printf("\n\nRunning Time:%dms\n",endtime-begintime);
}

本机运行结果：

0.09261000 0.00161630 0.99999869
0.13863000 0.00241948 0.99999707
0.08108000 0.00141507 0.99999900
0.17009000 0.00296854 0.99999559
0.17172000 0.00299699 0.99999551
0.12679000 0.00221284 0.99999755


Running Time:164844ms

F:\tcc\tcc\汇报文档1>

本机运行时间是164844ms，，，和十万次相比，整体运行速度快了

云机运行结果：

647999 0.00761771 0.99997098
0.90952998 0.01587316 0.99987401
0.89455998 0.01561192 0.99987813
0.72215003 0.01260319 0.99992058
0.82911003 0.01446977 0.99989531
0.04087000 0.00071329 0.99999975
0.02526000 0.00044086 0.99999990
0.84438002 0.01473624 0.99989142
0.57853001 0.01009678 0.99994903


Running Time:2520000ms
(base) [root@VM_0_17_centos 汇报文档1]#

云服务器运行时间2520000ms

一千万次不同浮点数的正弦值代码:

#include<stdio.h>
#include<math.h>
#include<stdlib.h>
#include<time.h> 
#define PI 3.1415
void main(){
	int begintime,endtime;
    int i = 0;
    float n;
	begintime=clock();
	for(i=0;i<10000000;i++)
	{
	n=(rand()%100000)*0.00001;
    printf("%.8f %.8f %.8f\n",n ,sin(n*PI/180),cos(n*PI/180));
}
    endtime = clock();
	printf("\n\nRunning Time:%dms\n",endtime-begintime);
}

本机运行结果：

0.22149999 0.00386578 0.99999253
0.13259000 0.00231406 0.99999732
0.27779001 0.00484819 0.99998825


Running Time:1228715ms

F:\tcc\tcc\汇报文档1>

本机运行时间是1228715ms，相对于一百万次又在整体上有所加速

云机运行结果：

0.06584000 0.00114909 0.99999934
0.69124001 0.01206377 0.99992723
0.64973998 0.01133952 0.99993571
0.78039998 0.01361973 0.99990725
0.19524001 0.00340747 0.99999419
0.97907001 0.01708666 0.99985401
0.87031001 0.01518874 0.99988464


Running Time:25350000ms

python

我使用的是python自带的编辑器idle，编译是在win10窗口下运行的。

十万次不同浮点数的正弦值代码

import numpy as np
import time

start=time.perf_counter()

t =np.arange(0,1,0.00001)
for i in t:
   s = np.sin(np.pi*i/180)
   c = np.cos(np.pi*i/180)
   print(i,s,c)
   
end=time.process_time()
print('Runing time.%s Seconds'%(end-start))

0.9999929999999999 0.01745228428284329 0.9998476972885975
0.9999939999999999 0.017452301733477622 0.9998476969839974
0.999995 0.01745231918411195 0.9998476966793972
0.999996 0.017452336634746275 0.9998476963747966
0.9999969999999999 0.017452354085380588 0.9998476960701957
0.9999979999999999 0.017452371536014904 0.9998476957655945
0.999999 0.017452388986649213 0.9998476954609931
Runing time.9.7488590785 Seconds

I:\python代码\汇报文档>

本机运行时间9.7488590785s ,和c语言相比慢好多

0.9999500000000001 0.017451533905561922 0.9998477103861083
0.9999600000000001 0.017451708411907303 0.9998477073402258
0.9999700000000001 0.017451882918252153 0.9998477042943128
0.9999800000000001 0.01745205742459647 0.9998477012483694
0.99999 0.017452231930940258 0.9998476982023956
Runing time.-15415.782753022 Seconds
(base) [root@VM_0_17_centos 汇报文档]#

一百万次不同浮点数的正弦值代码

import numpy as np
import time

start=time.perf_counter()

t =np.arange(0,1,0.000001)
for i in t:
   s = np.sin(np.pi*i/180)
   c = np.cos(np.pi*i/180)
   print(i,s,c)
   
end=time.process_time()
print('Runing time.%s Seconds'%(end-start))

0.9999992 0.01745239247677607 0.9998476954000727
0.9999992999999999 0.0174523942218395 0.9998476953696125
0.9999994 0.017452395966902932 0.9998476953391524
0.9999994999999999 0.01745239771196636 0.9998476953086922
0.9999996 0.017452399457029793 0.999847695278232
0.9999996999999999 0.01745240120209322 0.9998476952477718
0.9999998 0.017452402947156654 0.9998476952173116
0.9999998999999999 0.01745240469222008 0.9998476951868515
Runing time.67.41105205082 Seconds

I:\python代码\汇报文档>

0.999999 0.017452388986649213 0.9998476954609931
0.9999990999999999 0.01745239073171264 0.9998476954305329
0.9999992 0.01745239247677607 0.9998476954000727
0.9999992999999999 0.0174523942218395 0.9998476953696125
0.9999994 0.017452395966902932 0.9998476953391524
0.9999994999999999 0.01745239771196636 0.9998476953086922
0.9999996 0.017452399457029793 0.999847695278232
0.9999996999999999 0.01745240120209322 0.9998476952477718
0.9999998 0.017452402947156654 0.9998476952173116
0.9999998999999999 0.01745240469222008 0.9998476951868515
Runing time.-154181.281969009 Seconds
(base) [root@VM_0_17_centos 汇报文档]# ^C
(base) [root@VM_0_17_centos 汇报文档]#

一千万次不同浮点数的正弦值代码

import numpy as np
import time

start=time.perf_counter()

t =np.arange(0,1,0.0000001)
for i in t:
   s = np.sin(np.pi*i/180)
   c = np.cos(np.pi*i/180)
   print(i,s,c)
   
end=time.process_time()
print('Runing time.%s Seconds'%(end-start))

0.9999994 0.017452395966902932 0.9998476953391524
0.9999994999999999 0.01745239771196636 0.9998476953086922
0.9999996 0.017452399457029793 0.999847695278232
0.9999996999999999 0.01745240120209322 0.9998476952477718
0.9999998 0.017452402947156654 0.9998476952173116
0.9999998999999999 0.01745240469222008 0.9998476951868515
Runing time.463.665149412 Seconds

D:\python代码\汇报文档>

0.9999994999999999 0.01745239771196636 0.9998476953086922
0.9999996 0.017452399457029793 0.999847695278232
0.9999996999999999 0.01745240120209322 0.9998476952477718
0.9999998 0.017452402947156654 0.9998476952173116
0.9999998999999999 0.01745240469222008 0.9998476951868515
Runing time.-243078.302655293 Seconds

Java

我使用的是editplus3编辑器，编译是在win10窗口下运行的

十万次不同浮点数的正弦值代码

百万次不同浮点数的正弦值代码

千万次不同浮点数的正弦值代码

golang

我采用的编辑器是中国开发的liteide编辑器，也可以编译运行，这里我就全部在win10命令窗口运行

十万次不同浮点数的正弦值代码

package main //文件所在包main

import (
	"fmt"
	"math"
	"math/rand"
	"time"
	//引入包fmt
)

func main() { //func是一个关键字，main是一个主函数
	start := time.Now()
	for i := 0; i < 100000; i++ {
		c := rand.Float64()
		var pi float64 = 3.1415
		b := math.Sin(c * pi / 180)
		a := math.Cos(c * pi / 180)
		fmt.Println(a, b, c) //调用fmt这个包的一个函数
	}
	cost := time.Since(start)
	fmt.Printf("cost=[%s]", cost)
}

0.999994833390935 0.003214528182522622 0.18418464725410097
0.9999945974968972 0.0032870924870582663 0.18834242024015024
0.9998795833509061 0.015518337475975993 0.8891971577425763
0.9999194089985363 0.012695491641446928 0.7274390851448634
0.9999905713825652 0.004342481545227153 0.2488139852828219
0.9999478351489565 0.010214058004306184 0.5852498520869169
0.9999144314319856 0.013081659453180629 0.7495673616436112
0.9999535215359796 0.009641305295090791 0.5524309534935193
cost=[14.7315232s]
I:\golong程序\go_code>

0.9999905713825652 0.004342481545227153 0.2488139852828219
0.9999478351489565 0.010214058004306184 0.5852498520869169
0.9999144314319856 0.013081659453180629 0.7495673616436112
0.9999535215359796 0.009641305295090791 0.5524309534935193
cost=[1m31.476300684s](base) [root@VM_0_17_centos 汇报文档3]#

百万次不同浮点数的正弦值代码

package main //文件所在包main

import (
	"fmt"
	"math"
	"math/rand"
	"time"
	//引入包fmt
)

func main() { //func是一个关键字，main是一个主函数
	start := time.Now()
	for i := 0; i < 1000000; i++ {
		c := rand.Float64()
		var pi float64 = 3.1415
		b := math.Sin(c * pi / 180)
		a := math.Cos(c * pi / 180)
		fmt.Println(a, b, c) //调用fmt这个包的一个函数
	}
	cost := time.Since(start)
	fmt.Printf("cost=[%s]", cost)
}

0.9999924490061552 0.0038861202595829136 0.22266541695828487
0.9999373144212858 0.011196750776302633 0.6415588901093644
0.9999646985255648 0.008402481935495023 0.48144661653365883
0.9998784742980534 0.015589632304733274 0.8932826652358626
0.9999999997749794 2.121417230956505e-05 0.0012155183880337823
0.9999271682440302 0.012068894210938128 0.6915338849055698
cost=[1m19.1541632s]
I:\golong程序\go_code>

本机运行时间lm19.1541632s

0.9999646985255648 0.008402481935495023 0.48144661653365883
0.9998784742980534 0.015589632304733274 0.8932826652358626
0.9999999997749794 2.121417230956505e-05 0.0012155183880337823
0.9999271682440302 0.012068894210938128 0.6915338849055698
cost=[14m3.920909476s]

千万次不同浮点数的正弦值代码

package main //文件所在包main

import (
	"fmt"
	"math"
	"math/rand"
	"time"
) //引入包fmt,math,time,math/rand

func main() { //func是一个关键字，main是一个主函数
	start := time.Now()
	for i := 0; i < 10000000; i++ {
		c := rand.Float64()
		var pi float64 = 3.1415
		b := math.Sin(c * pi / 180)
		a := math.Cos(c * pi / 180)
		fmt.Println(a, b, c) //调用fmt这个包的一个函数
	}
	cost := time.Since(start)
	fmt.Printf("cost=[%s]", cost)
}

0.9998998223976465 0.014154333935401484 0.8110345976812745
0.9999838322843707 0.005686402189747993 0.3258182111597408
0.9998589817008006 0.016793353216026163 0.9622618739795078
0.9999988461621413 0.0015191031519480086 0.08704079978310163
0.9998551839929443 0.01701796234675305 0.975133248251753
0.9999997330416314 0.000730696014803921 0.04186703624658029
0.9999975365775767 0.0022196483455909797 0.1271803375083845
0.9999342300914827 0.0114688923333294 0.6571529141078967
cost=[7m41.5104498s]
I:\golong程序\go_code>

0.9998589817008006 0.016793353216026163 0.9622618739795078
0.9999988461621413 0.0015191031519480086 0.08704079978310163
0.9998551839929443 0.01701796234675305 0.975133248251753
0.9999997330416314 0.000730696014803921 0.04186703624658029
0.9999975365775767 0.0022196483455909797 0.1271803375083845
0.9999342300914827 0.0114688923333294 0.6571529141078967
cost=[35m18.423696655s]

octave

本机上我用的是最新版的octave5-1-0版本

十万次不同浮点数的正弦值代码

octave:16> tic;
octave:17> for k=1:100000
> a=rand(1,1);
> sin(a*pi/180);
> cos(a*pi/180);
> end
octave:18> toc;

 >> toc;
Elapsed time is 75.861 seconds.

octave:39> toc;
Elapsed time is 45.8987 seconds.

百万次不同浮点数的正弦值代码

octave:16> tic;
octave:17> for k=1:1000000
> a=rand(1,1);
> sin(a*pi/180);
> cos(a*pi/180);
> end
octave:18> toc;

>> toc;
Elapsed time is 119.613 seconds.

octave:30> toc;
Elapsed time is 86.1641 seconds.

千万次不同浮点数的正弦值代码

octave:16> tic;
octave:17> for k=1:10000000
> a=rand(1,1);
> sin(a*pi/180);
> cos(a*pi/180);
> end
octave:18> toc;

᐀>> toc;
Elapsed time is 575.841 seconds.

octave:33> toc;
Elapsed time is 296.095 seconds.

julia

云机使用的版本是julia1.1.1

十万次不同浮点数的正弦值代码

using Pkg
using PyCall
@pyimport math # 表示调用python 中math库
@pyimport datetime  #表示调用时间库
t1 = datetime.datetime.now() #计时开始
for i in 1:100000
    a=rand(1)
    b=math.sin(math.pi*a/180)
    c=math.cos(math.pi*a/180)
    println(b,c)
end
t2 = datetime.datetime.now()

println((t2 - t1))

本机运行结果时间

julia> t2 = datetime.datetime.now()
2019-10-28T20:30:22.456

julia> println((t2 - t1))
60125 milliseconds

云机运行结果时间

0.012042535624649922
0.008174816574793015

julia> t2 = datetime.datetime.now()
2019-10-28T14:27:40.169

julia> println((t2 - t1))
80715 milliseconds

百万次不同浮点数的正弦值代码

using Pkg
using PyCall
@pyimport math # 表示调用python 中math库
@pyimport datetime  #表示调用时间库
t1 = datetime.datetime.now() #计时开始
for i in 1:1000000
    a=rand(1)
    b=math.sin(math.pi*a/180)
    c=math.cos(math.pi*a/180)
    println(b,c)
end
t2 = datetime.datetime.now()

println((t2 - t1))

本机运行结果时间

julia> t2 = datetime.datetime.now()
2019-10-28T20:40:01.983

julia> println((t2 - t1))
287567 milliseconds

julia>

云机运行结果时间

julia> t2 = datetime.datetime.now()
2019-10-28T14:44:25.904

julia> println((t2 - t1))
346389 milliseconds

julia>

千万次不同浮点数的正弦值代码

using Pkg
using PyCall
@pyimport math # 表示调用python 中math库
@pyimport datetime  #表示调用时间库
t1 = datetime.datetime.now() #计时开始
for i in 1:10000000
    a=rand(1)
    b=math.sin(math.pi*a/180)
    c=math.cos(math.pi*a/180)
    println(b,c)
end
t2 = datetime.datetime.now()

println((t2 - t1))

本机运行结果时间

julia> t2 = datetime.datetime.now()
2019-10-28T20:48:52.046

julia> println((t2 - t1))
1305821 milliseconds

云机运行结果时间

julia> t2 = datetime.datetime.now()
2019-10-28T15:01:43.405

julia> println((t2 - t1))
1406510 milliseconds

总结：

本机

次数/语言	c语言	python	java	golang	octave	julia
十万次	31168ms	9.7488590785s	75448ms	14.7315232s	75.861 s	60125ms
百万次	164844ms	67.41105205082s	391668ms	lm19.1541632s	119.613 s	287567 ms
千万次	1228715ms	463.665149412 S	2002814ms	7m41.5104498s	575.841 s	1305821 ms

次数/语言	c语言	python	Java	golang	octave	julia
十万次	250000ms	-15415.782753022 S	9268ms	1m31.476300684s	45.8987s	80715ms
百万次	2520000ms	-15418.1281969009s	103415ms	14m3.920909476s	86.1641 s	346389 ms
千万次	25350000ms	-243078.302655293S	1576909ms	35m18.423696655s	296.095 s	1406510ms

具体结论我无法得出，由于电脑配置原因，运行速度有很大误差。各位可以试试