FOR loop calculating next values using the previous values

Hi there,

I have a pre-defined matrix 257 x 3 matrix, say Xin = [X Y Z]. Based on that, I would like to calculate a new matrix Xout = [M N P], using the values from Xin, such that e.g.:

M1 = (X2-X1) * sin(Y1*Z1) * cos(Y2*Z2)

M2 = (X3-X2) * cos(Y2*Z2) * sin (Y3*Z3)

and so on until the last value of 257.

So far I used the FOR loop (1:size(Xin,1)) and tried to pull the values from Xin using (i) and (i-1), however it states that "Subscript indices must either be real positive integers or logicals" that obviously happens when i=1.

I tried to overcome this by pre-defining X1,Y1,Z1 from Xin such e.g., X1 = Xin (1,1) and then start the FOR loop from (2:size(Xin,1)) but then it gives an error that index exceeds matrix dimensions.

X1,Y1,Z1 cannot initially be set up as some random values, since they come from the real survey data.

Any help would be appreciated. Thank you in advance!

[Jan]

Please post your code. Then it is easy to suggest a modification.

[John Pickles]

I posted in the code in the answer given by Aquatris, can you please have a look at that?

Answer 1

[Aquatris]

Your for loop cannot go upto size(Xin,1). Imagine Xin has 10 elements. For calculation of M10, you would need to access Xin(11), which does not exist. This is why you were getting index exceed matrix dimensions error.

For the "Subscript indices must either be real positive integers or logicals" error, I think you did not use "i" as your loop counter variable. "i" is by default equal to sqrt(-1), an imaginary number, in Matlab.

From what I understand you are trying to do;

X = Xin(:,1); Y = Xin(:,2); Z = Xin(:,3);
for i = 1:(size(Xin,1)-1)
    % assuming the 'M1=(X2-X1)*sin(Y1*Z1)*cos(Y2*Z2)' is the correct formula
    M(i) = (X(i+1)-X(i)) * sin(Y(i)*Z(i)) * cos(Y(i+1)*Z(i+1));    
end

[John Pickles]

Thanks for clarifying! Indeed, matrix dimension error occurs because it cannot calculate from Xin (11), as you mentioned (forgot to write this point down in the question).

After all, your code helped me to calculate the new values. However, since (i) is specified to range from 1:(size-1), it now also creates the same size 257x3 matrix, but the last row is zeros (as to overcome the problem above).

The code does calculate well and at this stage of the project that should be fine, however, is there a way to overcome this again and get the values for the last 257th survey station.

Here's the code I wrote down in the end (where New is a predefined 257x3 matrix):

GeoCoordinates = zeros (size(New)); 
    
MDn = New (:,1); 
INCLnew = New (:,2);
AZIMnew = New (:,3);

for i = 1:(size(New,1)-1)
    NS = ((MDn(i+1)-MDn(i))/2)*((sin(INCLnew(i))*cos(AZIMnew(i))+(sin(INCLnew(i+1)*cos(AZIMnew(i+1))))));
    
    EW = ((MDn(i+1)-MDn(i))/2)*((sin(INCLnew(i))*sin(AZIMnew(i))+(sin(INCLnew(i+1)*sin(AZIMnew(i+1))))));
    
    TVD = ((MDn(i+1)-MDn(i))/2)*(cos(INCLnew(i)+cos(INCLnew(i+1))));
    
    GeoCoordinates(i,:) = [NS, EW, TVD];
end 

Thanks!

[Aquatris]

I don't think so since the value of the last station requires information for the 258th values, which do not exist. You might want to check the origin of the equations and see how this is handled.