Tic Toc Function Help

Okay so I had a code which had an error, and I thought I fixed it, but now I'm getting another error in it

function [m1,m2]=mycomplexityplot(fh2,x2,N)

time1=zeros(1,N);

for i = 1:N
    tic
    for k1=1:i
        fh2(x2);
    end
    time1(i)=toc;
end

time2=zeros(1,N);
for i2=1:N
    tic
    for k2=1:i2
        for k3=1:i2
            fh2(x2);
        end
    end
    time2(i2)=toc;
end
n=1:N;

loglog(n,time1,'b-')
hold on

loglog(n,time2,'g--')
legend('1 loop','2 loops')
xlabel('N')
ylabel('time(s)')
hold off

p1=polyfit(n(length(n)/2:end),time1(n/2:end),1);
p2=polyfit(n(length(n)/2:end),time2(n/2:end),1);
m1=p1(1);
m2=p2(1);

when i try to run: fh2=@sqrt; [m1,m2]=mycomplexityplot(fh2,5,100)

I get the errors:

Error using polyfit (line 48) X and Y vectors must be the same size.

Error in mycomplexityplot (line 35) p1=polyfit(n(length(n)/2:end),time1(n/2:end),1);

Error in TestCasesLab9 (line 2) [m1,m2]=mycomplexityplot(fh2,5,100

what can do to fix this? sorry for the repost!!

Answer 1

[Matt Fig]

Try these lines instead:

p1=polyfit(n(length(n)/2:end),time1(length(n)/2:end),1);
p2=polyfit(n(length(n)/2:end),time2(length(n)/2:end),1);

[Matt]

thanks, Matt, it's actually running! But I am getting,

m1 =

   2.2743e-06

m2 =

   4.3520e-04

instead of

m1= 1

m2= 2

is there something wrong i'm doing in my code?

[Matt Fig]

Why do you expect those values?

[Matt]

Here is the actual function description, sorry, I should have included that in the first place:

mycomplexityplot, which takes in a function handle, fh2, a scalar value, x2, at which to evaluate fh2, and the number of iterations, N, up to which the for-loops should be evaluated. Use loglog to plot the time it takes to evaluate the function either inside one loop or inside two for-loops (each loop iterates N times) for increasing values of N. Finally, assuming the loglog plots are approximately linear, have the function output the slope of each line, rounded to the nearest whole number using polyfit. For the single loop calculation, call the slope m1. For the double loop calculation, call it m2.

[Matt Fig]

You still have not answered why you expect m1 to be 1 and m2 to be 2.....

[Matt Fig]

Perhaps you are looking for this:

p1=polyfit(log10(n(length(n)/2:end)),log10(time1(length(n)/2:end)),1);
p2=polyfit(log10(n(length(n)/2:end)),log10(time2(length(n)/2:end)),1);

If your question was answered, please accept the answer...

[Matt]

It is 1 and 2 because those are the numbers my instructor gave me to test if my code was right and there's something wrong with it. Thanks though