How to get kse or ksdensity as definable or anonymous function of x

From what I can tell both kse and ksdensity are used to get a smoothly defined probability distribution function from vector of finite data. Is there a way to get MATLAB to produce the actual function as the output instead of finite set of input/output vectors? I need to be able to call on the kernal density estimator at an arbitrary point in my domain and I'd prefer not to just interpolate between the values provided. Is there a functionality of either of these two that gives the true function/estimator or do I need to write my own code to accomplish this?

[Torsten]

What kind of "function" do you expect to get from discrete data you supply ?

You don't specify the form of the distribution function you want to fit your data to. So you only get discrete values back.

[David Gillcrist]

Either I have a misunderstanding about kernal density estimation, or I don't quite understand your response. Is it not the case that the result of producing a KSE (at least using a Gaussian smoothing function which is the default for kse) is a smoothly defined function?

[Torsten]

You can define the function as

f = @(y) ksdensity(x,y)

e.g.

rng('default')  % For reproducibility
x = [randn(30,1); 5+randn(30,1)];
f = @(y)ksdensity(x,y);
fy = f(1.34)

fy = 0.1094

Of course, the density function will always be recomputed this way.