21 views (last 30 days)

Show older comments

Hello,

I am getting confused with the basic concept of pointer in mex function. So I have couple of questions...

Below is an example from mathwork.

Q1.

As far as I know, in C++, when we declare double *y for the first time, y is the variable that stores address for data type double.

void arrayProduct(double x, double *y, double *z, mwSize n)

z[i] = x * y[i];

But I was wondering if z, y stores addresses not values how can we calculate like this z[i] = x * y[i]; ?

Shouldn't this be something like (*z)[i] and (*y)[i]? but then I see the message

error: subscripted value is neither array nor pointer nor vector

(*z)[i] = x * (*y)[i];

So I guess y stores the value.

Q2.

In the gateway function,

double *inMatrix; inMatrix is supposed to be a variable that stores addresses of data type double

inMatrix = mxGetPr(prhs[1]); inMatrix stores addresses of second argument of the input variable

Am I understanding this concept correctly?

Q3.

Then, 'inMatrix' is a pointer to 'A'

`y' is a pointer to 'inMatrix'

'multiplier' is a copy of 's'

'x' is a copy of 'multiplier'

Any help would be appreciated.

Thanks in advance.

#include "mex.h"

/* The computational routine */

void arrayProduct(double x, double *y, double *z, mwSize n)

{

mwSize i;

/* multiply each element y by x */

for (i=0; i<n; i++) {

z[i] = x * y[i];

}

}

/* The gateway function */

void mexFunction( int nlhs, mxArray *plhs[],

int nrhs, const mxArray *prhs[])

{

double multiplier; /* input scalar */

double *inMatrix; /* 1xN input matrix */

size_t ncols; /* size of matrix */

double *outMatrix; /* output matrix */

/* get the value of the scalar input */

multiplier = mxGetScalar(prhs[0]);

/* create a pointer to the real data in the input matrix */

inMatrix = mxGetPr(prhs[1]);

/* get dimensions of the input matrix */

ncols = mxGetN(prhs[1]);

/* create the output matrix */

plhs[0] = mxCreateDoubleMatrix(1,(mwSize)ncols,mxREAL);

/* get a pointer to the real data in the output matrix */

outMatrix = mxGetPr(plhs[0]);

/* call the computational routine */

arrayProduct(multiplier,inMatrix,outMatrix,(mwSize)ncols);

}

clc

clear all;

clear mex;

mex arrayProduct.c

s = 5;

A = [1.5, 2, 9];

B = arrayProduct(s,A)

James Tursa
on 17 Feb 2021

Probably the thing to do is brush up on C/C++ pointers. Maybe look through some online tutorials. E.g.,

double *z

means

z is of type "pointer to double"

z[i] is of type "double"

*z is of type "double"

and

(*z)[i]

makes no sense because you can't deference a double ... you can only dereference pointers to something.

Jan
on 17 Feb 2021

Q1. But I was wondering if z, y stores addresses not values how can we calculate like this z[i] = x * y[i]; ?

This is a question about C, not Matlab.

The [ ] operator in C means accessing the contents the pointer points to. So this is equivalent:

z[i]

*(z + i)

The first notation is just a nicer abbreviation.

Q2. inMatrix = mxGetPr(prhs[1]); inMatrix stores addresses of second argument of the input variable

Almost correct. mxGetPr replies the address of the data section of this Matlab variable. The address of the variable itself is the pointer in prhs[1].

Matlab variables contain a header, in which e.g. the dimensions of the data are stored, and a pointer to an array, which contains the actual values.

Q3. I do not understand, what is the question here. I see some correct statements.

James Tursa
on 17 Feb 2021

In fact, the compiler turns all expressions of the form z[i] into *(z+i) for compilation. And since this pointer addition is commutative, that means all of the following are equivalent expressions as far as the compiler is concerned:

z[5]

*(z+5)

*(5+z)

5[z]

Yes, as odd as it looks, that last one is valid C syntax and evaluates the same as the other three expressions, because it is turned into the 3rd expression for compilation.

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!