# Kernel density estimation Problem graph ?

3 views (last 30 days)
Lakshit Chugh on 3 May 2020
Answered: Shubh Sahu on 6 May 2020
l = linspace(-4,4,500);
fhat = zeros(size(l));
h = 1.06*n^(-1/5);
hold on
for i=1:n % get each kernel function evaluated at x% centered at data
f = exp(-(1/(2*h^2))*(l-x(i)).^2)/sqrt(2*pi)/h;
plot(l,f/(n*h));
fhat = fhat+f/(n);
end
plot(l,fhat);
hold off
Thats my code for kernel density estimation where x(i) is dataset of mixture of gaussian and beta distribution but the graph i am getting is value so small that even i cant see it and i am not sure is it right?

Shubh Sahu on 6 May 2020
See this piece of code:
f = exp(-(1/(2*h^2))*(l-x(i)).^2)/sqrt(2*pi)/h;
plot(l,f/(n*h));
In this f is a point and l is a vector. So in first place you are plotting point against a vector so it will be very difficult to see. Make fhat an array.
fhat(i)= fhat(i-1)+f/(n);
Note: I will strongly recommend you to see this link. Never plot anything inside for loop until it's very necessary to do that.