Fast calculation of stable density functions based on off-line precomputations

This sample code implements a proposal to compute arbitrary stable densities based on interpolation.

---------------------------------------------------------------
Fast calculation of stable density functions based on off-line
precomputations. Application to ML parameter estimation
---------------------------------------------------------------

Federico Simmross-Wattenberg, Marcos Martín-Fernández, Pablo
Casaseca-de-la-Higuera and Carlos Alberola-López

[e-mail]

ETSI Telecomunicación.
Universidad de Valladolid.
Campus Miguel Delibes
47011 Valladolid
Spain

This sample code implements a proposal to compute arbitrary stable densities
based on interpolation over a precomputed grid. Included files are:

stablepdf_interp.m: returns an arbitrary stable probability density function.

stablefit.m: estimates stable parameters for a given data sample.

stablefit_init.m: returns a quick guess of stable parameters for a given data
sample.

stable_table_30dB_30dB_1e8.mat: a precomputed table of stable densities, ready
to use with stablepdf_interp.m and stablefit.m. It covers 0.4<=alpha<=2,
-1<=beta<=1 and -1e8<=x<=1e8.

stabletabulate.m: computes the grid to be used with stablepdf_interp.m
and stablefit.m. It needs a handle to a function that calculates arbitrary
stable PDF points (not included, please see help for references). Note that the
function the handle points to has to behave well over the whole requested
parameter space or the grid construction will fail.

stable_adap_sample.m, compute_ab_grid.m: these functions are internal to
stabletabulate.m. Please use it instead.