// Initialize the filter using some model
void pf_init_model(pf_t *pf, pf_init_model_fn_t init_fn, void *init_data)
{
int i;
pf_sample_set_t *set;
pf_sample_t *sample;
set = pf->sets + pf->current_set;
// Create the kd tree for adaptive sampling
pf_kdtree_clear(set->kdtree);
set->sample_count = pf->max_samples;
// Compute the new sample poses
for (i = 0; i < set->sample_count; i++)
{
sample = set->samples + i;
sample->weight = 1.0 / pf->max_samples;
sample->pose = (*init_fn) (init_data);
// Add sample to histogram
pf_kdtree_insert(set->kdtree, sample->pose, sample->weight);
}
pf->w_slow = pf->w_fast = 0.0;
// Re-compute cluster statistics
pf_cluster_stats(pf, set);
//set converged to 0
pf_init_converged(pf);
return;
}
// Initialize the filter using a guassian
void pf_init(pf_t *pf, pf_vector_t mean, pf_matrix_t cov)
{
int i;
pf_sample_set_t *set;
pf_sample_t *sample;
pf_pdf_gaussian_t *pdf;
set = pf->sets + pf->current_set;
// Create the kd tree for adaptive sampling
pf_kdtree_clear(set->kdtree);
set->sample_count = pf->max_samples;
pdf = pf_pdf_gaussian_alloc(mean, cov);
// Compute the new sample poses
for (i = 0; i < set->sample_count; i++)
{
sample = set->samples + i;
sample->weight = 1.0 / pf->max_samples;
sample->pose = pf_pdf_gaussian_sample(pdf);
// Add sample to histogram
pf_kdtree_insert(set->kdtree, sample->pose, sample->weight);
}
pf->w_slow = pf->w_fast = 0.0;
pf_pdf_gaussian_free(pdf);
// Re-compute cluster statistics
pf_cluster_stats(pf, set);
//set converged to 0
pf_init_converged(pf);
//allow converged to be 1 if we actually start out converged
pf_update_converged(pf);
return;
}
// Create a new filter
pf_t *pf_alloc(int min_samples, int max_samples,
double alpha_slow, double alpha_fast,
pf_init_model_fn_t random_pose_fn, void *random_pose_data)
{
int i, j;
pf_t *pf;
pf_sample_set_t *set;
pf_sample_t *sample;
srand48(time(NULL));
pf = calloc(1, sizeof(pf_t));
pf->random_pose_fn = random_pose_fn;
pf->random_pose_data = random_pose_data;
pf->min_samples = min_samples;
pf->max_samples = max_samples;
// Control parameters for the population size calculation. [err] is
// the max error between the true distribution and the estimated
// distribution. [z] is the upper standard normal quantile for (1 -
// p), where p is the probability that the error on the estimated
// distrubition will be less than [err].
pf->pop_err = 0.01;
pf->pop_z = 3;
pf->dist_threshold = 0.5;
pf->current_set = 0;
for (j = 0; j < 2; j++)
{
set = pf->sets + j;
set->sample_count = max_samples;
set->samples = calloc(max_samples, sizeof(pf_sample_t));
for (i = 0; i < set->sample_count; i++)
{
sample = set->samples + i;
sample->pose.v[0] = 0.0;
sample->pose.v[1] = 0.0;
sample->pose.v[2] = 0.0;
sample->weight = 1.0 / max_samples;
}
// HACK: is 3 times max_samples enough?
set->kdtree = pf_kdtree_alloc(3 * max_samples);
set->cluster_count = 0;
set->cluster_max_count = max_samples;
set->clusters = calloc(set->cluster_max_count, sizeof(pf_cluster_t));
set->mean = pf_vector_zero();
set->cov = pf_matrix_zero();
}
pf->w_slow = 0.0;
pf->w_fast = 0.0;
pf->alpha_slow = alpha_slow;
pf->alpha_fast = alpha_fast;
//set converged to 0
pf_init_converged(pf);
return pf;
}