Capturing outputs from Loops

Hi

I have produced some code but now need to look at how it performs under various changes in input. This (I hope!) is best done by using loops. Capturing the various outputs and plotting them is where I am stuck!

for n=1:0.1:10 
% Various Functions
error1=
error2=
end

error1 and error2 are produced and work well in isolation. I need to vary n and capture error1 and error2 for each n. Plotting error1 and error2 against n would be the end target.

I can't quite see how to do this!

Any help in how to do this would be very welcome. How to measure and capture the tic/toc for each n would be useful as well.

Cheers

Joe

Answer 1

[Alexandra Harkai]

To keep track of error1/2 and times, you can maintain an array for each of them:

error1 = zeros(91,1);
error2 = zeros(91,1);
time = zeros(91,1);
for n=1:0.1:10 
tic;
% Various Functions
m = n*10-9;
t(m) = toc;
error1(m)= <first error>;
error2(m)= <second error>;
end

plot([error1, error2, t]);

Since n is not always an integer, m=n*10-9 gives an integer indexing.

[Steven Lord]

Rather than multiply a decimal number to get an index (which does NOT always work) consider iterating over a vector of integer values and transforming the elements of that vector into a decimal.

for n = 1:0.1:5
  n2 = 10*n-9;
  disp([n, n2])
  if n2 ~= round(n2)
     fprintf('10*(%g)-9 returns %d%+g.\n', n, round(n2), n2-round(n2));
  end
end

Compare with:

for n2 = 1:41
  n = (n2+9)/10;
  disp([n, n2])
  if n2 ~= round(n2)
     fprintf('10*(%g)-9 returns %d%+g.\n', n, round(n2), n2-round(n2));
  end
end