Conversion of C++ code to matlab code

I've a C++ code to be converted to MATLAB script. Following is the code. Please assist in the conversion process. I've tried my part in conversion which I've attached

#include <iostream>
#include <math>
#include <complex>
#include <bits/stdc++.h>
#include <bits>
using namespace std;
void main()
{
const k=10;
const Rmn=12;
float phi=0;
float theta=90;
float phimn=14;
float thetamn=16;
int q=34;
const dx=23;
const dy=25;
int m,n;
cout<<"Enter the value of M"
cin>>m
cout<<"Enter the value of N"
cin>>n
int FinalAF=0;

for(int i=-m+1;i<=m;i++)
{
    for(int j=-n+1;j<=n;j++)
    {
       
        // e1=sin(-1*k*Rmn)+jcos(-1*k*Rmn)
        // e2=cos(phimn)+jsin(phimn)
       
        complex<double> compl1(0,-1*k*Rmn);
        complex<double> compl2(0,phimn);
        Inter=exp(compl1)*exp(compl2)/Rmn;
        Imn=Inter*cos(thetamn)^q;
       
        float xmn=(m-0.5)*dx;
        float ymn=(n-0.5)*dy;
       
        string rmn = to_string(xmn)+"x"+to_string(ymn)+"y"
      
        string rc = to_string(sin(theta)*cos(phi))+"x"+to_string(sin(theta)*cos(phi))+"y"+to_string(sin(theta)*cos(phi))+"z"
       
        complex<double> compl3(0, rmn*rc*k);
        AF=Imn*exp(compl3);
       
        FinalAF=FinalAF+AF;
         
    }
}

int Etheta = FinalAF*cos(theta);   
}

% The code that I tried converting from C++ to MATLAB

%k=10;
%rmn=12;
%phi=0;
%theta=90;
%thetamn=16;
%phimn=14;
%q=34;
%dx=23;
f=input('Enter the frequency in GHz');
rmn=input('Enter the distance form feed phase center to “mnth” element');
phimn=input('Enter the phase provided by the mnth element of microstrip reflectarray array');
q=input('Enter q for evaluating the cosine power radiation pattern');
epsilonr=2.2;
c = 3*1e8;
lambda0 = c/(f*1e9);
lambdag = lambda0/sqrt(epsilonr);
dx = lambdag/2;
k = 2*pi/lambda0;
M=input('Enter the number of rows');
N=input('Enter the number of columns');
string r = string(sin(theta)*cos(phi))+"x"+ string(sin(theta)*cos(phi))+"y"+ string(sin(theta)*cos(phi))+"z";
for m= -M+1:M
    for n= -N+1:N
        g=phimn-k*rmn;
        Imn=(exp(1i*g)*(cos(phimn)^q))/rmn;
    end
end
for m = -(M+1):M
    for n = -(N+1):N
AF = Imn*exp(1i*(rmn.*r)*k); 
% .* = dot product of rmn vector and r vector
    end
end

% let radiation pattern be E
for m = -(M+1):M
    for n = -(N+1):N
		E = AF*cos(phimn);
    end
end

[KSSV]

You have converted thats fine...what is the problem now?

[Venkat Raman]

Since I'm new to MATLAB, I didn't know how to assign/initialize vectors with their actual meanings. The above code attempts to plot the 3D radiation pattern of an antenna array. For plotting the 3D radiation pattern, I've read tutorials saying to use 'surf'/'plot3' function. Three variables are to be inputted to the surf function.

These are the set of equations

[Walter Roberson]

        string rmn = to_string(xmn)+"x"+to_string(ymn)+"y"
      
        string rc = to_string(sin(theta)*cos(phi))+"x"+to_string(sin(theta)*cos(phi))+"y"+to_string(sin(theta)*cos(phi))+"z"
       
        complex<double> compl3(0, rmn*rc*k);

What is the intended meaning of using the * operator between strings ? I do not find any definition for such an operation in C++ references. It seems difficult to believe that the result would be numeric suitable for use... whatever compl3 is.

[Venkat Raman]

Those lines of code indicate the vector r

rcap is given by the equation below. When I tried this in MATLAB, it threw an error stating string is obsolete and insisted to use char. x,y and z are unit normal vectors of the coordinate system

The * attempts to multiply two trigonometric components. I think there is something to be fixed here.

[Walter Roberson]

You are trying to multiply STRINGS. You are not creating two three-D vectors and multiplying them: you are calculating the numeric portions of the vectors and converting them to text and adding in the characters 'x' and 'y' and 'z' in appropriate places, and then you are trying to multiply the text !

[Guillaume]

edit: I need my eyes tested, there's indeed a multiplication between strings in the C++ code which is completely meaningless. The C++ code couldn't have worked.

If I were given the task to convert code from one language to matlab, I would do it like this:

• Understand what the code does
• Write down the algorithm/equation implemented by said code
• throw away that code and never look at it again
• implement the algorithm/equation from step 2 in matlab

I would certainly not attempt to convert the code line by line. Attempting to write matlab code as if it was C++ is going to lead to very poor code.

As it appear you already have the equations, I would start from there and forget about the C++ code.

The equations can be implemented very easily in matlab without any loop. This will make it much easier to read as well. However, it's not clear to me why has a vector arrow above it nor what , , and are.

Answer 1

[Bhaskar R]

You have not used loop iteration varibales in C++ code?, anyway the plane conversion of your code is

% using namespace std;
% void main()
% {
% const k=10;
k = 10;
% const Rmn=12;
Rmn=12;
% float phi=0;
phi=0;
% float theta=90;
theta=90;
% float phimn=14;
phimn=14;
% float thetamn=16;
thetamn=16;
% int q=34;
q=34;
% const dx=23;
dx=23;
% const dy=25;
dy=25;
% int m,n;
% cout<<"Enter the value of M"
% cin>>m
m = input('Enter the value of M : ');
% cout<<"Enter the value of N"
% cin>>n
n = input('Enter the value of N : ');
% int FinalAF=0;
FinalAF=0;
% for(int i=-m+1;i<=m;i++)
% {
for ii = -m+1:m
    %     for(int j=-n+1;j<=n;j++)
    %     {
    for jj = -n+1:n
        %  // e1=sin(-1*k*Rmn)+jcos(-1*k*Rmn)
        % // e2=cos(phimn)+jsin(phimn)
        % complex<double> compl1(0,-1*k*Rmn);
        compl1 = 0+(-1*k*Rmn)*1i;
        % complex<double> compl2(0,phimn);
        compl2 = 0+(phimn)*1i;      
        %Inter=exp(compl1)*exp(compl2)/Rmn;
        Inter = exp(compl1)*exp(compl2)/Rmn;
        % Imn=Inter*cos(thetamn)^q;
        Imn=Inter*cos(thetamn)^q;
        % float xmn=(m-0.5)*dx;
        xmn=(m-0.5)*dx;
        %         float ymn=(n-0.5)*dy;
        ymn=(n-0.5)*dy;
        %  string rmn = to_string(xmn)+"x"+to_string(ymn)+"y"
        rmn = [num2str(xmn), 'x', num2str(ymn), 'y'];
        % string rc = to_string(sin(theta)*cos(phi))+"x"+to_string(sin(theta)*cos(phi))+"y"+to_string(sin(theta)*cos(phi))+"z"
        rc = [num2str(sin(theta)*cos(phi)), 'x',...
            num2str(sin(theta)*cos(phi)), 'y', num2str(sin(theta)*cos(phi)), 'z'];
        %  complex<double> compl3(0, rmn*rc*k);
        compl3 = 0+ (rmn*rc*k)*1i;
        %AF=Imn*exp(compl3);
        AF=Imn*exp(compl3);
        % FinalAF=FinalAF+AF;
        FinalAF=FinalAF+AF;
    end
end
% int Etheta = FinalAF*cos(theta);
Etheta = FinalAF*cos(theta);

[Venkat Raman]

I went through your code. It was really helpful. Since I'm new to MATLAB, I didn't know how to assign/initialize vectors with their actual meanings. The above code attempts to plot the 3D radiation pattern of an antenna array. For plotting the 3D radiation pattern, I've read tutorials saying to use 'surf'/'plot3' function. Three variables are to be inputted to the surf function.

The set of equations are:

How to proceed with the surf function?

[Walter Roberson]

Which variable is to be on the x axes? Which variable is to be on the y axes? Which variable is to be on the z axes?

[Venkat Raman]

The electric field which is a function of theta, phi and the distance vector is along the z axis, phi is along x axis and theta along y axis. phi and theta are measured in degrees while E field is measured in V/m

Answer 2

[rania craise]

please is there any way to convert the program from a C++ language to matlab ?

[Walter Roberson]

Probably only by hand.