Help troubleshooting iteration code

Hi,

I have an iteration code which I am using to find latitude/longitude of a set of heights (h_intercept). This is a 1x79 matrix.

It works perfectly until the 22nd value. I've found that this is when h_test>h_intercept. I tried to put a condition in to reset it but it doesn't work.

Furthermore, the strange thing is that the iteration runs and finds the height - the issue is that it is taking the wrong height.

For example

   h_intercept=sat_look_pass1_llh(3,60:79)/2e3;
for j=1:length(h_intercept)
rng_sat= sat_look_tcs_pass1(3,j); 
  u_sat=[sat_look_tcs_pass1(1,j)/sat_look_tcs_pass1(3,j);sat_look_tcs_pass1(2,j)/sat_look_tcs_pass1(3,j);sat_look_tcs_pass1(3,j)/sat_look_tcs_pass1(3,j)];
h_intercept=sat_look_pass1_llh(3,j)/2e3;
  h_test=zeros(1,3);
rng_test_min=0;
rng_test_max=rng_sat/2e3;
err=0.01;
while abs(h_intercept-h_test)>err   
    rng_test=(rng_test_min+rng_test_max)/2;
    tcs_test=u_sat*rng_test;
    llh_test=tcs2llhT(tcs_test,station_llh);
    h_test=llh_test(3,:);
     if h_test>=h_intercept
         rng_test_max=rng_test; 

    else
        rng_test_min=rng_test;
    end

end copter_llh(:,j)=(llh_test); h_interceptloop(:,j)=(h_intercept); end % code end

This code is using values 60:79 of my matrix and although it runs, it is actually finding h_intercept(1,1:19)

It seems that after h_test>h_intercept, it starts from the beginning of the height matrix. How is that possible?!

Any suggestions appreciated!

[Buzz]

Well, I made a silly mistake with the for loop. Should be for i=59:79. However, the condiiton for h_test>h_intercept still doesn't work

[Andy L]

are you saying that the code should read

 for i = 59:79

instead of

 for j = 1:length(h_intercept)

[Buzz]

Yeah, I realise that it was giving the wrong height because I had for j=1:length(h_intercept). So it does work for h_test<h_intercept but not when it is more

Answer 1

[Image Analyst]

Here's a virtually foolproof way to solve it http://blogs.mathworks.com/videos/2012/07/03/debugging-in-matlab/

[Buzz]

Thanks!

Answer 2

[Guillaume]

You start with this line:

   h_intercept=sat_look_pass1_llh(3,60:79)/2e3;

which as you say uses columns 60 to 79 of your sat_look_pass1_llh matrix. However, in the for loop, you overwrite it with:

h_intercept=sat_look_pass1_llh(3,j)/2e3;

with j from 1 to 19 (length(h_intercept)). Thus you're iterating over columns 1 to 19. As it is the first line serves no purpose.

I suspect you wanted to do:

h_intercepts=sat_look_pass1_llh(3,60:79)/2e3;
for j = 1:length(h_intercepts)
  h_intercept = h_intercepts(j);
  ...
end

which you could also simplify with

h_intercepts=sat_look_pass1_llh(3,60:79)/2e3;
for h_intercept = h_intercepts
  ...
end

[Buzz]

Yeah, thanks! Silly mistake. Still can't fix the iteration for h_test>h_intercept, though

[Andy L]

you set h_test to 1x3 zero's with this line

 h_test = zeros(1,3);

Are you expecting h_intercept to be negative? Either way in your while loop you then overwrite h_test with

 h_test = llh_test(3,:);

Have you checked what it is that these variables contain.

h_test is an array and h_intercept is 1x1 - is this correct?

You need to use the debugger as image analyst suggests to check your variables are doing what you expect them to do. I tend to work out the loop by hand first to check what it is I should be getting (certainly when it is relatively simple like this one anyway).

Keyboard is one way to get started with debugging.

[Buzz]

h_test=zeros(1,3) doesn't change anything. I was testing it out and can just be h_test=0, as an initial condition. Yeah, I stepped through it and I know that the values that don't work are greater than h_test (474m). I know that my if condition is not working, as it should step back

[Buzz]

when h_test>h_intercept, all range values become zero