Index exceeds the number of array elements (1). Error coming in the code below

Hi,

I am using one of the matlab function called CircleFitByTaubin but I am stuck on one of the line as shown in the error below and not sure how to correct this? If anyone can help here? I have attached the code, function and data here.

Error:

Index exceeds the number of array elements (1).

Error in CircleFitByTaubin (line 25)

Yi = XY(i,2) - centroid(2); % centering data

Error in circlefitting (line 9)

Par = CircleFitByTaubin(P);

[Mathieu NOE]

ok

can you share your main code ?

[muhammad choudhry]

Hi, main code is already attached if u see in the comment above named bestcirclefit.m and for function circlefitbytaubin.m and I also posted text1.txt file with data

[per isakson]

But you have not told us how to use these three files to reproduce the error, which you have reported!

[muhammad choudhry]

Hi,

there was 2 script attached one was best circle fit in which I am calling the function CircleFitByTaubin, which is also attached above please see CircleFitByTaubin.m , I want to use this code to cover most of the points in the circle and estimate the size of that circle and radius.

Thanks

Answer 1

[per isakson]

I cannot reproduce (on R2018b) the error that you report:

Index exceeds the number of array elements (1).
Error in CircleFitByTaubin (line 25)
    Yi = XY(i,2) - centroid(2);  %  centering data
Error in circlefitting (line 9)
Par = CircleFitByTaubin(P);

You have not provided the mfunction, circlefitting. There is no call to CircleFitByTaubin in bestcirclefit.m. The mfile, bestcirclefit.m, has an syntax error and isn't possible to run.

CircleFitByTaubin works as expected:

%%
XY = dlmread('test1.txt','\t',0,0); 
%%
par = CircleFitByTaubin(XY);

and returns

>> par
par =
      0.41541      0.30058     0.028124

[Mathieu NOE]

Hi

after some clean up - both functions are now working fine

test scripts provided in attachement

[muhammad choudhry]

Thanks for the clean up, can you tell me what did you do differently, I also want to know is there any possibility to cover most of the point in the circle as currently you can see not many of the points are on the circle.

Please see the attach!

Regards,

M Choudhry

[Mathieu NOE]

well , the CircleFitByTaubin works as expected:.

the other file (bestcirclefit.m) had the wrong name and was also inconsistent with the supposed name of that function (circlefit.m)

there were some minor stuff like removing unnecessary code

If you edit the files before / after my corrections , you should be able to see what I did

regarding you're experimental data, they are what they are even though they are a bit off from a circle (or the part that you measured has been squeezd ?)

there is nothing that you can do about that... otherwise you want to fit an ellipse ?

[muhammad choudhry]

thanks alot for the reply, how to do the ellipse fitting. The major goal is to fit the points to the well defined shape like circle, if you can guide me how to do with the ellipse it will give me a plus point to show in my report that how most of the points are covered through the ellipse.

Thanks alot once again.

[per isakson]

Search the File Exchange for fit ellipse. There are some submissions, which will help you earn the "plus point".

"to fit the points to the well defined shape" it's the other way round. One fits a model/equation/shape to data points.

And in the column to the right of this page there is an interesting link: Fitting Circles (again) and Ellipses