plot a smooth curve

Hello everyone,

I plotted a curve with a set of data by using a spline technique, but the problem is the start and the end of the curve is matching with the original dataset. My code is listed below. Is there any tecnhique or command that makes the curve more smoother or can we use the fitting technique. please help me in that.

Thanks

clc

clear

[x]=[1.111409702 2.561607677 4.152025868 5.825439427 7.677672228 9.669974952 11.68269125 13.63751636 15.45892018 17.11664862 18.82457281];

[y]=[0.05 0.3 0.74996023 9.99803011 13.68801508 14.57288221 11.85807424 4.94788522 2.04185776 0.49090234 0.03087319];

plot (x, y, 'g', 'LineWidth',2);

grid

A = trapz(x, y)

Increase = 10;

newX = linspace(1,20, 25 * Increase);

smoothedY = spline(x, y, newX);

hold on

% plot(newXSamplePoints, smoothedY, '-ob');

plot(newX, smoothedY);

hold off

Answer 1

[Kunal Kandhari]

Hi Mohanned,

It sounds like you want a kind of interpolation, because "smoothing" usually trims the values of the extreme points of a curve, whereas interpolation fits those points exactly (as per your requirement that "the peak should be at same point").

I have done it using pchip.

clc
clear
[x]=[1.111409702	2.561607677	4.152025868	5.825439427	7.677672228	9.669974952	11.68269125	13.63751636	15.45892018	17.11664862	18.82457281];
[y]=[0.05	0.3	0.74996023	9.99803011	13.68801508	14.57288221	11.85807424	4.94788522	2.04185776	0.49090234	0.03087319];
plot (x, y, 'g', 'LineWidth',2);
grid
A = trapz(x, y)
Increase = 10;
newX = linspace(1,20, 25 * Increase);
smoothedY = spline(x, y, newX);
% hold on
% % plot(newXSamplePoints, smoothedY, '-ob');
% plot(newX, smoothedY);
% hold off

Xi = 0:0.1:19;
Yi = pchip(x,y,Xi);
hold on
plot(Xi,Yi)
hold off

There are other 1-D data interpolation methods to choose from, go to method — Interpolation method section in this page, so you should take a look and pick one that best suits your needs.

Hope this helps!

Answer 2

[Nikhil Sapre]

You can use the curveFitter app.

Curve Fitting - MATLAB & Simulink (mathworks.com)