«

Hi,

Does anyone know how to make the density plot, more specifically a
gaussian-smoothed version of a histogram. Actually, it's not exactly
a plot that I want. What I want to do is get the maximum count (or
highest peak) of the density distribution. It's different from the
probability density function that has the center at zero.

In S plus and R, there is the "density" function which I can use to
get the max x and y in the frequency distribution.

I really appreciate if anyone can help.

On Thu, Dec 11, 2008 at 1:57 PM, <> wrote:
> Hi,
>
> Does anyone know how to make the density plot, more specifically a
> gaussian-smoothed version of a histogram. Actually, it's not exactly
> a plot that I want. What I want to do is get the maximum count (or
> highest peak) of the density distribution. It's different from the
> probability density function that has the center at zero.
>
> In S plus and R, there is the "density" function which I can use to
> get the max x and y in the frequency distribution.

scipy ML is more appropriate for this kind of questions:

http://www.scipy.org/Mailing_Lists

In scipy, there is a kernel density estimator, which is one popular
method for non-parametric density estimation, and fit exactly the
"Gaussian smoothed version of the histogram" when used with Gaussian
kernels.

David