gbkreg - Kernel non linear regression function

**gbkreg** [

*options*]

Kernel estimation of conditional moments. Data are read from standard input as
couple (x,y). The moments of y are computed on a regular grid in x. The kernel
bandwidth, if not provided with the option

**-H**, is set automatically.

**-n**
- number of equispaced points where moments are computed
(default 64)

**-H**
- set the kernel bandwidth

**-S**
- scale the automatic kernel bandwidth

**-K**
- choose the kernel to use (default 0)

- 0
- Epanenchnikov

- 1
- Rectangular

- 2
- Gaussian

**-M**
- choose the method to compute the density (default 1)

- 0
- FFT (number of points rounded to nearest power of 2)

- 1
- discrete convolution (only with compact kernels)

- 2
- explicit summation (can be long)

**-O**
- set the output, comma separated list of m mean, v standard
deviation, s skewness and k kurtosis (default m)

**-v**
- verbose mode

**-F**
- specify the input fields separators (default "
\t")

**-h**
- this help

- gbkreg -M 2 < file
- compute the kernel regression of the entries in the second
column of 'file' vs. the entries in the first column. If more data columns
exist in file they are ignored. Explicit summation method (slower) is
used.

- gbkreg -02 < file
- compute the kernel regression of the standard deviation of
the entries in the second column of 'file' vs. the entries in the first
colum

Written by Giulio Bottazzi

Report bugs to <gbutils@googlegroups.com>

Package home page
<http://cafim.sssup.it/~giulio/software/gbutils/index.html>

Copyright © 2001-2015 Giulio Bottazzi This program is free software; you
can redistribute it and/or modify it under the terms of the GNU General Public
License (version 2) as published by the Free Software Foundation;

This program is distributed in the hope that it will be useful, but WITHOUT ANY
WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR
A PARTICULAR PURPOSE. See the GNU General Public License for more
details.