How to find the width of a peak where there is no actual data, just individual points?

Question

Alfredo Scigliani 2023-6-27

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1988483-how-to-find-the-width-of-a-peak-where-there-is-no-actual-data-just-individual-points

评论： Alfredo Scigliani 2023-6-27

I am trying to obtain the width at the half maximum location of the peak displayed in the figure. I have already found the location of the half max and have depicted it with dotted lines. Now, I would like to calculate the width inside that peak at the specified height. I would just get the intercept of the blue line and the dotted line and calculate it but since there is no actual data in the blue line, just a straight line connecting the individual points, I am not sure how to do it.

clear; clc; clf; close all;

xdata = [0 0.012 0.03 0.051 0.08 0.11 0.15];

ydata = [0.060583 0.098032 0.1407 0.40137 0.1407 0.098032 0.060583];

halfMax_y = 0.20069;

maxLocation_x = 0.051;

figure;

plot(xdata, ydata, 'o-');

hold on;

yline(halfMax_y, 'k--');

xline(maxLocation_x, 'k--');

I would like to find this distance, you can assume that the units of the x axis are distance in meters:

If somebody has an idea on how to do it I would really appreciate it. And if it can also be kept as simple as possible because I am a beginner I appreciate it. I have seen a couple peak processing toolbox online but I am not experienced with those options.

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

David Goodmanson 2023-6-27

1
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1988483-how-to-find-the-width-of-a-peak-where-there-is-no-actual-data-just-individual-points#answer_1263193

在 MATLAB Online 中打开

Hi Alfredo,

Usually your ydata would be the frequency domain version of a linear quantity such as voltage or displacement. Half maximum refers to half of a squared quantity such as power, so it would be down from the peak by a factor of sqrt(2) for the linear quantity. I will arbitrarily assume that here, but if you really intend to drop by a factor of 2 in ydata you can make the appropriate change below.

With this small number of points it seems best to use some spline-type interpolation, which I did with pchip.

xdata = [0 0.012 0.03 0.051 0.08 0.11 0.15];
ydata = [0.060583 0.098032 0.1407 0.40137 0.1407 0.098032 0.060583];
npts = 1000;
xp = linspace(0,.15,npts);
yp = pchip(xdata,ydata,xp);
[ymax,indmax] = max(yp) 
halfh = ymax/sqrt(2);               % down by sqrt(2)
x1 = xp(1:indmax);
x2 = xp(indmax:end);
y1 = yp(1:indmax);
y2 = yp(indmax:end);
w1 = interp1(y1,x1,halfh);
w2 = interp1(y2,x2,halfh);
fwhm = w2-w1                       % full width half max
figure(1);
plot(xdata, ydata, 'o-',xp,yp);
hold on
plot([w1 w2],[halfh halfh],'og-')
hold off
grid minor
   ymax =  0.4014
   indmax = 341
   
   fwhm = 0.0246