Connecting Data points in a smooth curve

How can I connect 2D data points in a smooth curve given that I can't use spline because it requires my x elements to be unique?

[madhan ravi]

Could you upload the code?

[Aquatris]

You can average your data to have unique x-y pairs and use regular fits? Is there any reason why you cant average your data at same x values?

[Mohammad Farhat]

huge code and a lot of data points, but for simplicity the problem narrows down to having for example: x=[2 4 5 2 8 5]; y=[15 6 17 28 90 80]; if I spline my data, I get the following error: Error using chckxy (line 51) The data sites should be distinct. Error in spline (line 53) [x,y,sizey,endslopes] = chckxy(x,y);

So apparently, x elements should be unique.

[Mohammad Farhat]

Thanks Aquatris, but I can't average

Answer 1

[Akira Agata]

You can do this task by applying the method used in the 2nd example in spline function help page.

https://www.mathworks.com/help/matlab/ref/spline.html#bvjdts5

The following is an example:

% Your data
x = [2 4 5 2 8 5];
y = [15 6 17 28 90 80];

% Cubic spline data interpolation
t = 1:numel(x);
xy = [x;y];
pp = spline(t,xy);
tInterp = linspace(1,numel(x));
xyInterp = ppval(pp, tInterp);

% Show the result
figure
plot(x,y,'o:')
hold on
plot(xyInterp(1,:),xyInterp(2,:))
legend({'Original data','Spline interpolation'},'FontSize',12)

[Mohammad Farhat]

thanks that was helpful, turns out i have another problem and maybe spline isnt the answer at all because i guess it connects consecutive data points y_x and y_x+1, and I dont want this, so maybe the only solution is some graphic tool? is there any?

[Image Analyst]

This looks like a 1-D situation, not a 2-D situation. You have one input/independent/x value, and one output/dependent/y value for each input value. That's 1-D, even if they can be plotted in an x-y plane.

Attach your data, and show exactly what you'd like as the output. Maybe you just want interp1() but you just have to average the values that don't have unique x values. Or else add a tiny bit of noise to the x values to separate them.