How to find cumulative distribution function from a given data set(vector)

clc
clear all
rng('default')  % For reproducibility
y = evrnd(0,3,100,1);
[a,s]=cdfplot(y);

In this "cdfplot "directly plots the cdf graph, that too is repeats the each number then finds cdf.

But I want the cdf vector.

Can someone help me please?

Answer 1

[Paul]

Hi Shankul,

I'm not sure what "I want the cdf vector" means. If that means the data that forms the cdf plot on the graph, then the data can be obtained from the XData and YData properties of a:

rng('default')  % For reproducibility
y = evrnd(0,3,100,1);
[a,s]=cdfplot(y);

plot(a.XData,a.YData)

If that's not what you want, can the Question be clarified?

[Shankul Saini]

okay thanks @Paul but if you seem only 100 elements were there in y but in cdfplot we have 200 elements and in that each element is repeated once.

And also +inf and -inf are added in starting and last. How can i get plot which has only original number of elements which in this case is 100.

And one more thing, can i get this xdata and ydata without plotting, i don't need the plot i just need data.

[Paul]

Mabye ecdf is what you're looking for?

rng('default')  % For reproducibility
y = evrnd(0,3,100,1);
[a,s]=cdfplot(y);
[f,x]=ecdf(y);
hold on
plot(x,f)

whos x f y

  Name        Size            Bytes  Class     Attributes

  f         101x1               808  double              
  x         101x1               808  double              
  y         100x1               800  double              

The extra point in x and f is the first where x(1) = x(2) and f(1) = 0;

[Shankul Saini]

Okay yeah, that seems fine. Thanks paul