File: mese.sci

package info (click to toggle)
scilab 4.0-12
  • links: PTS
  • area: non-free
  • in suites: etch, etch-m68k
  • size: 100,640 kB
  • ctags: 57,333
  • sloc: ansic: 377,889; fortran: 242,862; xml: 179,819; tcl: 42,062; sh: 10,593; ml: 9,441; makefile: 4,377; cpp: 1,354; java: 621; csh: 260; yacc: 247; perl: 130; lex: 126; asm: 72; lisp: 30
file content (36 lines) | stat: -rw-r--r-- 842 bytes parent folder | download | duplicates (2) 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
 
function [sm,fr]=mese(x,npts);
//<sm,fr]=mese(x [,npts]);
//Calculate the maximum entropy spectral estimate of x
//  x    :Input sampled data sequence
//  npts :Optional parameter giving number of points of fr and sm
//        (default is 256)
//  sm   :Samples of spectral estimate on the frequency grid fr
//  fr   :npts equally spaced frequency samples in [0,.5)
//!
// author: C. Bunks  date: 24 August 1988
// revised: 9 April 1991
// Copyright INRIA

//default evaluation
 
   [rhs,lhs]=argn(0);
   if lhs==1 then,
      npts=256;
   end,
 
//estimate autocorrelation function of x
 
   Nx=length(x);
   r=convol(x,x(Nx:-1:1))
   r=r(Nx:-1:1)/Nx;
 
//get solution to the Yule-Walker equations
 
   [ar,sigma2,rc]=lev(r);
 
//compute spectrum
 
   ak=[1;ar];
   [sf,fr]=frmag(ak,npts);
   sm=sigma2*ones(sf)./(sf.*conj(sf));
endfunction