Ho to find the interception

I have a fitted Guassian plot to which I have plotted yline. Now I am interested to extrac two x- interception points through which my line is passing. I am not able to find any easy way to do it. so any help is appreciable.

code:

clc;close all;clear;

resultsdir = 'results';

Row = dir('cross_section_*.mat');

%Coefficients (with 95% confidence bounds):

a1 = 394.5 ;

b1 = 985.1 ;

c1 = 114.8 ;

a2 = 511.4 ;

b2 = 878.5 ;

c2 = 231.2 ;

a3 = -529.8 ;

b3 = 989.8 ;

c3 = 133.6 ;

xData = zeros(1,1640);

yData = zeros(1,1640);

color = ['r','g','b']

for X = 1:9

Filename = Row(X).name;

dest = fullfile(resultsdir, Filename);

Filedata(X) = load(Filename);

figure()

plot(Filedata(X).x)

[xData, yData] = prepareCurveData( [], Filedata(X).x);

ft = fittype( 'gauss3' );

opts = fitoptions( 'Method', 'NonlinearLeastSquares' );

opts.Display = 'Off';

opts.Lower = [-Inf -Inf 0 -Inf -Inf 0 -Inf -Inf 0];

opts.StartPoint = [446.913 842 37.5821195501848 401.619376603545 761 47.8742720232778 393.885950193204 926 55.7197242033307];

[fitresult, gof] = fit( xData, yData, ft, opts );

figure()

h = plot( fitresult,'k', xData, yData);

set(h,'LineWidth', 2)

title(sprintf('Name:%s , R-Square = %0.4f',Filename,gof.rsquare))

legend( h, 'Experiment', 'fit ', 'Location', 'NorthEast', 'Interpreter', 'none' );

% Label axes

ylabel('Intensity values')

xlabel( 'Along X of the CMOS', 'Interpreter', 'none' );

grid on

f = fitresult.a1*exp(-((xData-fitresult.b1)/fitresult.c1).^2) + fitresult.a2*exp(-((xData-fitresult.b2)/fitresult.c2).^2) + fitresult.a3*exp(-((xData-fitresult.b3)/fitresult.c3).^2);

Max(X) = max(f)

exp_perc(X) = Max(X)/exp(2);

yline(exp_perc(X))

% Max(X) = max(fitresult.val)

% f = a1*exp(-((Filedata(X)-b1)/c1).^2) + a2*exp(-((Filedata(X)-b2)/c2).^2) + a3*exp(-((Filedata(X)-b3)/c3).^2);

end

plot:

Answer 1

[Star Strider]

If you simply want to do it yourself, try something like this —

xData = 1:150;
yData = 50*exp(-(xData-75).^2 * 0.01) + randn(1,150);

ft = fittype( 'gauss3' );
opts = fitoptions( 'Method', 'NonlinearLeastSquares' );
[fitresult, gof] = fit( xData(:), yData(:), ft, opts );
h = plot( fitresult,'k', xData, yData);
ylineval = 10;                                                      % This Needs To Be Stated Separately From The 'yline' Call
yline(ylineval, 'DisplayName','yline');

hold on

xfit = h(2).XData;                                                  % Get Fitted Curve Data
yfit = h(2).YData;                                                  % Get Fitted Curve Data

ixv = find(diff(sign(yfit-ylineval)));                              % Calculate The Approximate Intersection Indicess

for k = 1:numel(ixv)
    idxrng = max(1,ixv(k)-1) : min(numel(xfit),ixv(k)+1);
    xv(k) = interp1(yfit(idxrng), xfit(idxrng), ylineval);          % Interpolate 'Exact' Values
    yv(k) = interp1(xfit(idxrng), yfit(idxrng), xv(k));
end
% xv
% yv
plot(xv, yv, 'sr','MarkerSize',10, 'DisplayName','Intersections')
text(10,40,sprintf('Intersections:\nX = %.3f\nX = %.3f',xv))

hold off

.

[Navaneeth Tejasvi Mysore Nagendra]

Thank you, this is extremely helpful

[Star Strider]

As always, my pleasure!

Answer 2

[KSSV]

The best you can use is: https://in.mathworks.com/matlabcentral/fileexchange/22441-curve-intersections

[Navaneeth Tejasvi Mysore Nagendra]

Thank you for the suggestion, I will take a look