Error using chckxy: The first input must contain unique values, while using only unique values

Hello,

So I have to spline some data I acquired with ginput. While using spline, I get the following error

Error using chckxy

The first input must contain unique values.

At first I thought I had accidently selected two or more equal points, so I tried selecting random points which I knew were different for sure, and I still get the same error.

Then, I get another error in spline itself:

[x,y,sizey,endslopes] = chckxy(x,y);

Keep in mind this is not even part of my code, this is just matlab's function.

The code I use to acquire the points is this one:

download clone embed print report

mov = VideoReader('pendulo.mp4'); % estrutura com video 
nFrames = mov.NumFrames; % numero de frames 
framerate = mov.FrameRate; % numero de frames por segundo 
dtframes=1/framerate; %tempo entre frames 
 
i=1; t=0; tf=nFrames*dtframes; dt=10*dtframes; 
if ~exist('dados.mat','file') 
 while (t <= tf) 
 mov.CurrentTime=t; frame=readFrame(mov); image(frame); drawnow 
 tv(i)=t; t=t+dt; i=i+1; title(strcat('Frame ',num2str(i))); 
 [x(i) ,y(i)]=ginput(1); 
end 
save ('dados.mat', 'x', 'y'); 
load ('dados.mat', 'x', 'y'); 
end

and the code I use for the spline function is this one:

load ('dados.mat', 'x', 'y'); 
 
xx = linspace(0,500,100); 
yy = spline(x,y,xx); 
 
figure; 
plot(x, y, 'ro'); 
hold on; 
plot(xx, yy, 'b', 'LineWidth', 1.5);

I don't understand why I get an error at MATLAB's own function, and why I get the "must contain unique values" error, even though they are unique.

The data is annexed as 'dados.mat'.

[dpb]

Post a complete working example here that illustrates the problem, not on some external site.

Need the data mostly, of course; nothing we can do without it to diagnose anything...

[Hélio Filho]

Pardon me, I used the pastebin because I also posted this on reddit.

I shall edit the main thread.

Answer 1

[dpb]

>> load dados
>> whos x y
  Name      Size            Bytes  Class     Attributes

  x         1x58              464  double              
  y         1x58              464  double              

>> numel(unique(x))
ans =
    36
>> numel(unique(y))
ans =
    57
>> 

Well, clearly your data are NOT unique...you've only got 36 x values and 57 (closer, anyways) y values.

>> xu
xu =
  Columns 1 through 18
         0   15.7431   16.7265   17.7099   18.6934   19.6768   20.6602   47.2127   68.8481   86.5497   87.5331   89.5000   91.4669   92.4503   94.4171   95.4006  103.2680  133.7541
  Columns 19 through 36
  162.2735  163.2569  166.2072  167.1906  168.1740  169.1575  170.1409  171.1243  173.0912  189.8094  210.4613  222.2624  224.2293  225.2127  226.1961  227.1796  228.1630  262.5829
>> histc(x,xu)
ans =
  Columns 1 through 30
     1     3     3     6     1     3     1     1     1     1     1     2     1     2     2     2     1     1     2     1     3     2     1     1     1     1     1     1     1     2
  Columns 31 through 36
     1     2     2     1     1     1
>> 

However you thought you generated unique values clearly didn't work as you presumed...

Above is scatter plot of the point cloud; the filled values are those that are unique in x; I didn't add y but the this shows the problem.

[dpb]

For such data, spline won't be the tool you're looking for anyways; you'll want to interpolate instead; unfortunately, interp1 and friends are also not adept at handling input vectors with duplicated values of the independent variable nor with the problem here of the multi-valued y values.

You may end up just needing to do pointwise lookup and local interpolation; I'm not thinking of any builtin MATLAB tools that are appropriate (but then again, maybe I'm just not thinking of something obvious).

[Hélio Filho]

Unfortunately my assignment requires me to use the spline tool. I have since given up on this, and am working with less points. I asked my professor and he told me to interpolate y with t so I don't get duplicate values. Not sure what he means by this though.

Even so, you've put a lot of work into this, thank you for the help!

[Hélio Filho]

So I've been thinking, maybe I should use the variable t as my x data and I should acquire the y variable from the ginput function. As this is a movie, t is constantly evolving, so I can guarantee that there will be no two t values that are the same.

However, while changing the acquiring to t, like this:

i=1; t=0; tf=nFrames*dtframes; dt=10*dtframes; %% 17 frames ao invés de 58
if ~exist('dados.mat','file')
 while (t <= tf)
 mov.CurrentTime=t; frame=readFrame(mov); image(frame); drawnow
 tv(i)=t; t=t+dt; i=i+1; title(strcat('Frame ',num2str(i)));
 [t(i) ,y(i)]=ginput(1);
end
save ('dados.mat', 't', 'y');
load ('dados.mat', 't', 'y');
end

It only allows me to take in 1 value; after that it automatically uses splice. I don't understand why, since i is evolving all the time inside the while, and therefore i should have t(i+1) all the time. Is t not representing time here?

Answer 2

[Hélio Filho]

So apparently I had two different points with the same X value, I had to decrease the number of points. This solution is not optimal, since I lose data, but if any of you have a better one I will take it!