# Matrix inversion and summation

4 views (last 30 days)
BeeTiaw on 29 Jun 2018
Commented: BeeTiaw on 30 Jun 2018
Hi all, can anyone help me to explain the mistake in my code below? The problem is as follow:
• The following parameter is assumed,
lma = 3;
ma = zeros(1,lma);
ma(2) = -0.2;
ma(lma) = -0.01;
• The parameter is then used to calculate the coordinate z using the following equation,
where z = x + i*y (complex number)
dumZ = 0;
sumdumZ = 0;
theta = linspace(0,2*pi,37);
xi = exp(1i.*theta);
for ii=1:length(theta)
for jj=1:lma
dumZ = ma(jj).*xi(ii).^jj;
sumdumZ = sumdumZ + dumZ;
end
zcon(ii) = 1./xi(ii)+sumdumZ;
dumZ = 0;
sumdumZ = 0;
end
zcon=zcon';
The following profile is obtained by plotting zcon,
• Now, using the hypothetical data above, i.e. the zcon, I want to back-calculate the variable "ma" or "alpha" as shown in the equation above. To do this, I first created a matrix A, which essentially a matrix containing all the constants. The following code is used to generate matrix A with 370 terms,
lx = length(zcon);
N = 10*37;
A = zeros(lx,N);
for ii=1:lx
for jj=1:N
A(ii,jj) = xi(ii).^jj;
end
end
A=[1./xi' A]; % insert a new colum to the left containing the 1/xi(ii) into the existing matrix A;
• Then the variable "ma" or "alpha" can be easily obtained using the following operation in Matlab,
maiter = A\zcon;
which give me a [370 x 1] matrix. To validate this, I again calculate the "predicted" z using the following code:
zdirect = A*maiter;
Which results an exactly the same curve (as expected) as shown below,
Now, coming to my question. Instead of using a direct calculation as shown above, i.e. the code "zdirect = A*maiter, instead I would like to manually use summation to calculate the "z" for each "theta" using the code below. Somehow, I could not get it right and cannot find the mistake. Below is the code that I used to calculate "z"
dumZ = 0;
sumdumZ = 0;
pow = [-1 1:1:N];
for ii=1:length(theta)
for jj=1:length(maiter)
dumZ = maiter(jj).*xi(ii).^pow(jj);
sumdumZ = sumdumZ + dumZ;
end
ziter(ii) = sumdumZ;
dumZ = 0;
sumdumZ = 0;
end
Instead, this is what I got,

David Goodmanson on 29 Jun 2018
Hi Bee,
Are you plotting the right variable? I ran your code, and zcon, zdirect and ziter all reproduce the same curve. Only the simple exponential variable xi gives the red circle in the second plot.

Show 1 older comment
BeeTiaw on 30 Jun 2018
@David Goodmanson, can you please copy-paste your result that in here? so that we can compare the real and imaginary part.
David Goodmanson on 30 Jun 2018
Hi Bee,
Below are the first few rows of [zcon zdirect ziter]. The only change I made to your code was
zcon = zcon.' instead of zcon = zcon'
so that the complex conjugate of zcon is not taken. That means that the table I get is the complex conjugate of yours. I also added
ziter = ziter.'
so that ziter is a column vector. Of course neither of these changes/additions affects the shape of the plot, 'triangle' vs. circle.
0.7900 + 0.0000i 0.7900 + 0.0000i 0.7900 + 0.0000i
0.7882 - 0.2471i 0.7882 - 0.2471i 0.7882 - 0.2471i
0.7815 - 0.4792i 0.7815 - 0.4792i 0.7815 - 0.4792i
0.7660 - 0.6832i 0.7660 - 0.6832i 0.7660 - 0.6832i
0.7363 - 0.8484i 0.7363 - 0.8484i 0.7363 - 0.8484i
0.6862 - 0.9680i 0.6862 - 0.9680i 0.6862 - 0.9680i
0.6100 - 1.0392i 0.6100 - 1.0392i 0.6100 - 1.0392i
0.5039 - 1.0633i 0.5039 - 1.0633i 0.5039 - 1.0633i
0.3666 - 1.0446i 0.3666 - 1.0446i 0.3666 - 1.0446i
0.2000 - 0.9900i 0.2000 - 0.9900i 0.2000 - 0.9900i
0.0093 - 0.9077i 0.0093 - 0.9077i 0.0093 - 0.9077i
-0.1975 - 0.8061i -0.1975 - 0.8061i -0.1975 - 0.8061i
-0.4100 - 0.6928i -0.4100 - 0.6928i -0.4100 - 0.6928i
-0.6167 - 0.5741i -0.6167 - 0.5741i -0.6167 - 0.5741i
-0.8058 - 0.4545i -0.8058 - 0.4545i -0.8058 - 0.4545i
BeeTiaw on 30 Jun 2018
Hi David Goodmanson,
I got it correct now after change it as per your suggestion. Thanks a lot!
It seems that it has something to do with consistency of the conjugate/non-conjugate of the complex number. Although I still do not clearly understand it.
But I would consider the problem is solved now.