Classes | Defines | Functions
contrib/brl/bbas/bsta/algo/bsta_sample_set.h File Reference

Classes to collect samples. More...

#include <bsta/bsta_parzen_sphere.h>
#include <bsta/bsta_mixture.h>
#include <bsta/bsta_attributes.h>
#include <bsta/bsta_gaussian_sphere.h>
#include <bsta/bsta_gaussian_full.h>
#include <vnl/vnl_vector_fixed.h>
#include <vnl/vnl_matrix_fixed.h>
#include <vcl_fstream.h>
#include <vcl_utility.h>
#include <vcl_iostream.h>

Go to the source code of this file.

Classes

class  bsta_sample_set< T, n >
 A class to hold samples, the window width parameter, weights for each sample, assignments of each sample to modes/cluster centers/classes. More...

Defines

#define MIN_VAR_   0.0001

Functions

template<class T >
bool bsta_sample_set_variance (const bsta_sample_set< T, 1 > &set, int mode, T min_var, T &out)
 compute the variance of a particular assignment in a bsta_sample_set.
template<class T , unsigned n>
bool bsta_sample_set_variance (const bsta_sample_set< T, n > &set, int mode, vnl_matrix_fixed< T, n, n > &out)
 compute the variance of a particular assignment in a bsta_sample_set.
template<class T , unsigned n>
bool bsta_sample_set_marginalize (const bsta_sample_set< T, n > &set, unsigned component, bsta_sample_set< T, 1 > &out_set)
 compute the marginalized 1D sample set distribution from nD set.
template<class T >
bool bsta_sample_set_fit_distribution (const bsta_sample_set< T, 1 > &set, bsta_mixture< bsta_num_obs< bsta_gaussian_sphere< T, 1 > > > &out)
template<class T , unsigned n>
bool bsta_sample_set_fit_distribution (const bsta_sample_set< T, n > &set, bsta_mixture< bsta_num_obs< bsta_gaussian_full< T, n > > > &out)
template<class T >
bsta_sample_set_log_likelihood (const bsta_sample_set< T, 1 > &set, bsta_num_obs< bsta_gaussian_sphere< T, 1 > > &dist, T total_weight)
 Total weight is used to normalize the weight of the distribution (bsta_num_obs class contains total weight of samples assigned to this distribution)
template<class T , unsigned n>
bsta_sample_set_log_likelihood (const bsta_sample_set< T, n > &set, bsta_num_obs< bsta_gaussian_full< T, n > > &dist, T total_weight)
 Total weight is used to normalize the weight of the distribution (bsta_num_obs class contains total weight of samples assigned to this distribution)
template<class T >
bsta_sample_set_log_likelihood (const bsta_sample_set< T, 2 > &set, bsta_num_obs< bsta_gaussian_sphere< T, 1 > > &dist0, T w0, bsta_num_obs< bsta_gaussian_sphere< T, 1 > > &dist1, T w1, T &w_sum)
 Total weight is used to normalize the weight of the distribution (bsta_num_obs class contains total weight of samples assigned to this distribution)
template<class T >
bool bsta_sample_set_print_to_m (const bsta_sample_set< T, 2 > &set, vcl_ofstream &of)
 a specialized matlab file printer for 2D data.
template<class T >
bool bsta_sample_set_dist_print_to_m (const bsta_sample_set< T, 2 > &set, vcl_ofstream &of)
 a specialized matlab file printer to visualize printed distribution.
template<class T >
bool bsta_sample_set_dist_print_to_m (const bsta_sample_set< T, 1 > &set, vcl_ofstream &of)
 a specialized matlab file printer to visualize printed distribution.

Detailed Description

Classes to collect samples.

Author:
Ozge C. Ozcanli
Date:
March 04, 2009
    Modifications
     (none yet)
   

Definition in file bsta_sample_set.h.


Define Documentation

#define MIN_VAR_   0.0001

Definition at line 27 of file bsta_sample_set.h.


Function Documentation

template<class T >
bool bsta_sample_set_dist_print_to_m ( const bsta_sample_set< T, 2 > &  set,
vcl_ofstream &  of 
)

a specialized matlab file printer to visualize printed distribution.

Definition at line 405 of file bsta_sample_set.h.

template<class T >
bool bsta_sample_set_dist_print_to_m ( const bsta_sample_set< T, 1 > &  set,
vcl_ofstream &  of 
)

a specialized matlab file printer to visualize printed distribution.

Definition at line 455 of file bsta_sample_set.h.

template<class T >
bool bsta_sample_set_fit_distribution ( const bsta_sample_set< T, 1 > &  set,
bsta_mixture< bsta_num_obs< bsta_gaussian_sphere< T, 1 > > > &  out 
)

Definition at line 194 of file bsta_sample_set.h.

template<class T , unsigned n>
bool bsta_sample_set_fit_distribution ( const bsta_sample_set< T, n > &  set,
bsta_mixture< bsta_num_obs< bsta_gaussian_full< T, n > > > &  out 
)

Definition at line 226 of file bsta_sample_set.h.

template<class T >
T bsta_sample_set_log_likelihood ( const bsta_sample_set< T, 1 > &  set,
bsta_num_obs< bsta_gaussian_sphere< T, 1 > > &  dist,
total_weight 
)

Total weight is used to normalize the weight of the distribution (bsta_num_obs class contains total weight of samples assigned to this distribution)

Definition at line 261 of file bsta_sample_set.h.

template<class T , unsigned n>
T bsta_sample_set_log_likelihood ( const bsta_sample_set< T, n > &  set,
bsta_num_obs< bsta_gaussian_full< T, n > > &  dist,
total_weight 
)

Total weight is used to normalize the weight of the distribution (bsta_num_obs class contains total weight of samples assigned to this distribution)

Definition at line 286 of file bsta_sample_set.h.

template<class T >
T bsta_sample_set_log_likelihood ( const bsta_sample_set< T, 2 > &  set,
bsta_num_obs< bsta_gaussian_sphere< T, 1 > > &  dist0,
w0,
bsta_num_obs< bsta_gaussian_sphere< T, 1 > > &  dist1,
w1,
T &  w_sum 
)

Total weight is used to normalize the weight of the distribution (bsta_num_obs class contains total weight of samples assigned to this distribution)

Definition at line 311 of file bsta_sample_set.h.

template<class T , unsigned n>
bool bsta_sample_set_marginalize ( const bsta_sample_set< T, n > &  set,
unsigned  component,
bsta_sample_set< T, 1 > &  out_set 
)

compute the marginalized 1D sample set distribution from nD set.

Definition at line 175 of file bsta_sample_set.h.

template<class T >
bool bsta_sample_set_print_to_m ( const bsta_sample_set< T, 2 > &  set,
vcl_ofstream &  of 
)

a specialized matlab file printer for 2D data.

Definition at line 364 of file bsta_sample_set.h.

template<class T >
bool bsta_sample_set_variance ( const bsta_sample_set< T, 1 > &  set,
int  mode,
min_var,
T &  out 
)

compute the variance of a particular assignment in a bsta_sample_set.

Definition at line 98 of file bsta_sample_set.h.

template<class T , unsigned n>
bool bsta_sample_set_variance ( const bsta_sample_set< T, n > &  set,
int  mode,
vnl_matrix_fixed< T, n, n > &  out 
)

compute the variance of a particular assignment in a bsta_sample_set.

Definition at line 137 of file bsta_sample_set.h.