[R] Local linear regression: where can I find kernels?

Philipp Kunze pkunze at gwdg.de
Wed Oct 13 08:21:06 CEST 2010


Hi there,

What I'm trying to do is to calculate the line coefficients of a local
linear regression model "by hand". I'm aware that there are many great
packages out there that calculate the local expectation E(y|x) with
local linear regression, but that's not what I need. I need the
coefficients of the line that was fitted locally. So, does anybody know
of a package that gets me those coefficients? If not, I would calculate
the values "by hand" with a local least square fit using kernel weights
on my data. To do so, I need a function that calculates kernel weights,
i.e. K_h(x-x_i). Which function can I use?

Any help would be greatly appreciated!

Best regards,
Philipp



More information about the R-help mailing list