MSE and RMSE of a matrix

I have a reference vector u of size 1x8. In this matrix the 1st four columns are voltage readings and last four columns are current readings.i.e.

u=[v1 v2 v3 v4 i1 i2 i3 i4];

Likewise I have another set of 100 vectors.The size of each of them is also 1x8. The nature of those vectors is also same like u i.e. 1st four columns are voltages and last four columns are currents in each of them. If I write those 100 vectors in a matrix called two. The size of the matrix two becomes 100x8. Now I find the difference between each of the row of matrix two and u. i.e.This gives me another matrix Error of size 100x8. i.e.

Error(nn,:)= abs(u(1,:)-two(nn,:));

Now I want to find the MSE and RMSE of this matrix Error and plot both the MSE and RMSE.

[Sadiq Akbar]

This is my code. Is this correct?

u=[1 1.5 2 2.5 150 20 60 110];
two=randi(100,8);
for nn=1:100
Error(nn,:)= abs(u(1,:)-two(nn,:));
Error4sn0(nn)=abs(Error(nn,:)).^2;  % Sqaured Error of Four sources
MSE4sn0(nn) = sum(Error4sn0(nn,:))/8;
end
MSE4sn0=MSE4sn0';
MSE4sn0=sort(MSE4sn0,'descend');
t=1:100;
semilogy(t,smooth(MSE4sn0),'linewidth',2)
xlabel('Runs')
ylabel('MSE')
title('MSE of All ')

Answer 1

[KSSV]

If yt and ye are the column vectors of theory and experimental values. You can get MSE, RMSE using:

e = (ye - yt)    % Errors
suqre_e = (ye - yt).^2   % Squared Error
MSE = mean((ye - yt).^2)   % Mean Squared Error
RMSE = sqrt(mean((ye - yt).^2));  % Root Mean Squared Error

[Sadiq Akbar]

Thank you very much KSSV for your prompt response. Here you have taken only column vectors y and yt. But in my problem I have a row vector u of size 1x8 and a matrix two of size 100x8. I want to find Error between u and each of row of matrix two. After that I want to find its MSE and RMSE.

[KSSV]

The same will work for matrices also.....

[Sadiq Akbar]

Thank you very much for your tip. But please check my above code and tell me whether its correct?

[KSSV]

u=[1 1.5 2 2.5 150 20 60 110];
two=randi(100,8);
[m,n] = size(two) ; 

Error = abs(u-two) ; 
square_Error = abs(u-two).^2 ; 
MSE = (u-two).^2/m ; 
RMSE = sqrt((u - two).^2/m);

[Sadiq Akbar]

Thank you very much KSSV for your help. But I want to draw the plot also. I was thinking that if I find the square of the absolute of each row , then find its mean value. so that will give me MSE of that row. Thus I will be having 100 such values of all rows. Then I want to draw plot of this MSE vs t, where t=1:100; So can I do like that?

[KSSV]

You can straight away use plot. You will get eight curves.

[Sadiq Akbar]

Thank you very much for your help. I used the plot command directly, but it gave me the following error:

Error using plot

Vectors must be the same length.

Error in MSERMSEbyKSSV (line 12)

plot(t,MSE,t,RMSE);

[KSSV]

Check the sizes of t , MSE.