Hi there,
I know some basics of matlab, though I'm new to it. I use it only when I need any plot.
Now, I have some data points (x,y), which I have attached here. This is from my model output.
I want to import this file into matlab and plot this as a 2D plot in X,Y. The most important thing is that I need a smooth curve as in the attached figure named "Want.jpg". If I normally plot these points in MS excel, what I'm getting is shown in the figure named "Getting.jpg" and I don't want a plot like this.
Can anyone help me. Thanks in advance.
Cheers.

 Réponse acceptée

Meg Noah
Meg Noah le 13 Jan 2020
Modifié(e) : Meg Noah le 13 Jan 2020

1 vote

First, I added a header to the xlsx file so readtable can be used properly.
myData = readtable('Data_help.xlsx');
figure()
plot(myData.x,myData.y,'+')
hold on;
ywant = 0:0.01:1;
P = polyfit(myData.y,myData.x,6);
xwant = polyval(P,ywant);
plot(xwant,ywant)
SamplePolyfit.png

5 commentaires

Meg Noah
Meg Noah le 13 Jan 2020
You can also polyfit the interpolated data - maybe it's a little nicer. There is a program called smooth that does a really good job, but I don't have that toolbox.
myData = readtable('Data_help.xlsx');
figure()
plot(myData.x,myData.y,'+')
hold on;
ytmp = 0:0.01:1;
xtmp = interp1(myData.y,myData.x,ytmp,'v5cubic')
P = polyfit(ytmp,xtmp,6);
ywant = 0:0.001:1;
xwant = polyval(P,ywant);
plot(xwant,ywant)
SamplePolyfit.png
Thanks very much Meg Noah. This is exactly what I wanted.
You mentioned about the 'smooth' program. I checked my toolbox and I have it there. In case if I want to use it, how should I use it?
Additonally, I need a small clarification about the interpolation; what is the meaning of 'v5cubic' there?
Thanks
Meg Noah
Meg Noah le 13 Jan 2020
'v5cubic' is one of many interp1 options that matlab offers. I usually use 'pchip' or 'spline'. Check the documentation.
The smooth operator can operate just on the data:
x_smoothed = smooth(x);
or you can overwrite the default sliding window length, with a number, say, 7, to have it smooth over 7 elements:
x_smoothed = smooth(x, 7);
There are some other features. It's a great algorithm. I think I have the paper somewhere, it was published in a now defunct computer journal/magazine.
Oh it's the 1D interpolation. I have been looking at the documentation of 'interp'. Thanks Meg Noah.
'Smooth' seems an interesting operator. I'll look for that paper as well. Thanks.
Hi Meg Noah, I' m trying the same thing with the attached data set. But ending up in a more curvy plot. I need a smooth plot similar, but better, than the attached figure. Thanks

Connectez-vous pour commenter.

Plus de réponses (0)

Catégories

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by