如果你是需要快速搭建一个matlab调c/c++环境,这篇文章可以参考
有了c代码,想在matlab里面调用,可以参考我这个模板
matlab调用代码:
clear all
close all
clc
input1 =1;
input2 =2;
[output1,output2] = mexfunction(input1,input2);
output1
output2
这里面强调两个概念
1、Matlab里面所有变量都是矩阵,包括单变量也是1*1的矩阵
2、Maltab矩阵按列优先访问,这个和fortran保持一致
Cpp代码:
#include "mex.h"
void c_func(double input1,double input2,double *output1,double *output2)
{
}
void mexFunction(int nlhs, mxArray *plhs[], int nrhs, const mxArray *prhs[])
{
double *pdata;
// mcnoidal(waveheight,waveperiod,0.02,waterdeepth ,0, datanum,1, waveheight,flapdeepth);
pdata=mxGetPr(prhs[0]);
double input1 = *pdata;
int M = mxGetM(prhs[0]);
int N = mxGetN(prhs[0]);
printf("%d * %d\n",M,N);
pdata=mxGetPr(prhs[1]);
double input2 = *pdata;
int m = 3;
int n = 2;
plhs[0] = mxCreateDoubleMatrix(m, n, mxREAL);
double *output1;
output1 = mxGetPr( plhs[0]);
//列优先排列
output1[0*m+0] = 1;
output1[1*m+0] = 2;
output1[0*m+1] = 3;
output1[1*m+1] = 4;
output1[0*m+2] = 5;
output1[1*m+2] = 6;
plhs[1] = mxCreateDoubleMatrix(1, 1, mxREAL);
double *output2;
output2 = mxGetPr( plhs[1]);
output2[0] = -1;
//c_func(input1,input2,output1,output2) ;
}
plhs 全称 parameters left matlab中的左侧输出值
prhs全称 parameters left matlab中的右侧输入值
plhs[0] plhs[1]是输出,需要多少输出变量,那么就在c中用mxCreateDoubleMatrix申请多少
所有的数字,数组,二位数字,到c这边都是一维数组,且按列优先访问。
输出如下
注三行两列的二维数组访问方式:
int m=3;
int n=2;
//内存列优先排列,但赋值按逐行赋值
output1[0*m+0] = 1;
output1[1*m+0] = 2;
output1[0*m+1] = 3;
output1[1*m+1] = 4;
output1[0*m+2] = 5;
output1[1*m+2] = 6;
//列优先排列,但按列赋值
output1[0*m+0] = 1;
output1[0*m+1] = 3;
output1[0*m+2] = 5;
output1[1*m+0] = 2;
output1[1*m+1] = 4;
output1[1*m+2] = 6;
//另:行优先排列,按行赋值
output1[0*n+0] = 1;
output1[0*n+1] = 2;
output1[1*n+0] = 3;
output1[1*n+1] = 4;
output1[2*n+0] = 5;
output1[2*n+1] = 6;
在matlab 使用mex 编译一下(预先需要mex setup)
mex mexFunction.cpp 即可编译,然后在malab直接调用。