File: gaussian.h

package info (click to toggle)
mldemos 0.5.1-3
  • links: PTS, VCS
  • area: main
  • in suites: jessie, jessie-kfreebsd
  • size: 32,224 kB
  • ctags: 46,525
  • sloc: cpp: 306,887; ansic: 167,718; ml: 126; sh: 109; makefile: 2
file content (101 lines) | stat: -rw-r--r-- 3,453 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
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
/************************************************************************/
/* This file is part of libfgmm.				        */
/* 								        */
/* libfgmm is free software: you can redistribute it and/or modify      */
/* it under the terms of the GNU Lesser General Public License as published by */
/* the Free Software Foundation, either version 3 of the License, or    */
/* (at your option) any later version.				        */
/* 								        */
/* libfgmm 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 Lesser General Public License for more details.			        */
/* 								        */
/* You should have received a copy of the GNU Lesser General Public License    */
/* along with libfgmm.  If not, see <http://www.gnu.org/licenses/>.     */
/* 								        */
/* Copyright 2010        LASA  - EPFL   http://lasa.epfl.ch             */
/*                                                                      */
/*       Florent D'halluin   <florent.dhalluin@epfl.ch>		        */
/************************************************************************/
#ifndef _GAUSSIAN_H_
#define _GAUSSIAN_H_

/* 
   FOR INTERNAL USE ONLY : this file is not part of the public API to 
   libfgmm. If you need access to model parameters, please use the 
   fgmm_get_* functions defined in fgmm.h 
*/ 

#include "smat.h"


/** One gaussian distribution */

struct gaussian{
  _fgmm_real prior; /* prior probability */
  int dim;  /* dimensionality */ 
  _fgmm_real * mean ;
  struct smat * covar; /* covariance matrix */ 
  struct smat * covar_cholesky; /* cache for cholesky decomp of covar */ 
  struct smat * icovar_cholesky; /* cholesky matrix with inverse diagonal */
  _fgmm_real nfactor; /* cache for 1. / determinant of covar */
};   

/** compute the probability density at vector value 
    
    value should be at the same dimension than g->dim */
_minline _fgmm_real gaussian_pdf(struct gaussian* g, const _fgmm_real* x)
{
  _fgmm_real dist2;
  _fgmm_real dist = smat_sesq(g->icovar_cholesky,g->mean,x);
  dist *= .5;
  dist2 =  expf(-dist)*g->nfactor;
  if(dist2 == 0) dist2 = FLT_MIN;
  return dist2;
}

/** alloc memory for the gaussian 
    and init it to zero with identity covariance matrix
*/
void gaussian_init(struct gaussian* g,int dim);
void gaussian_free(struct gaussian* g);

void invert_covar(struct gaussian* g);

void dump(struct gaussian* g);

/* draw one sample from the gaussian */
void gaussian_draw(struct gaussian* g, _fgmm_real * out);

/* get the projection of the gaussian on the given dimensions 
 * if result in NULL or wrong dimension .. is it (re) alloc'd */
void gaussian_get_subgauss(struct gaussian* g, struct gaussian* result,
			   int n_dim, int * dims);

#define ranf() ( (_fgmm_real) rand())/RAND_MAX

/** random sample from normal law ( mu = 0, sigma = 1. ) **/ 
_minline _fgmm_real randn_boxmuller( void )
{
  _fgmm_real x1, x2, w;
  do {
    x1 = 2.0 * ranf() - 1.0;
    x2 = 2.0 * ranf() - 1.0;
    w = x1 * x1 + x2 * x2;
  } while ( w >= 1.0 );

  w = sqrt( (-2.0 * log( w ) ) / w );
  x1 *= w;
  /* x2 *= w */   /* 2nd indpdt gaussian */
  return x1;
};



/** incremental mean/var update */
void gaussian_update(struct gaussian * g, 
		     const _fgmm_real * datapoint, 
		     _fgmm_real learning_rate);

#endif // _GAUSSIAN_H_