In fact, this is the final action in the Gaussian elimination protocol that we discussed in Part 2. If U is definitely an d n upper-triangular matrix, we know how to solve the linear program Ux b using back again substitution. Lsq_mean = params(2) lsq_std = params(3) % what you wantįigure hold on plot(xAxis,yourData. Gaussian Elimination Octave How To Solve The. =lsqcurvefit(gausFun,x0,xAxis,yourData,options) % meat and potatoes Options=optimset('Display','off') % avoid pesky messages from lsqcurvefit (optional) X0 = % parameters need to be in a single variable Height_est = max(fauxData)*rand mean_est = fauxMean*rand std_est=fauxStd*rand One is the program, the other one is the matrix that were going to use and the next three. % % Provide estimates for initial conditions (for lsqcurvefit) % % Octave Code For the correct development of this program you have to dowload the five attachments below. YourData = fauxData % replace with your actual distribution Xaxis = 1:length(faux) fauxData = histc(faux,xaxis) This is for fitting a Gaussian FUNCTION, if you just want to fit data to a Normal distribution, use 'normfit.'Ĭheck it % % Generate synthetic data (for example) % % %įaux = fauxStd.*randn(1,nPoints) + fauxMean % REPLACE WITH YOUR ACTUAL DATA I found that the MATLAB 'fit' function was slow, and used 'lsqcurvefit' with an inline Gaussian function. Which you can cross-check with x(3) = R (there should only be small differences). Which you then have to reverse-engineer to find the mean μ and the standard-deviation σ: mu = -x(2)/x(1)/2 The vector x you found this way will equal x = Now, solve for the linear system Ax=b with (these are Matlab statements): % design matrix for least squares fitĪ = If you know for sure your data y will be well-described by a Gaussian, and is reasonably well-distributed over your entire x-range, you can linearize the problem (these are equations, not statements): y = 1/(σ

You'll find ready-made implementations here, or here, or here for 2D, or here (if you have the statistics toolbox) (have you heard of Google? :)Īnyway, there might be a simpler solution. Fitting a single 1D Gaussian directly is a non-linear fitting problem.

Gaussian elimination octave