2 * folded-ctf is an implementation of the folded hierarchy of
3 * classifiers for object detection, developed by Francois Fleuret
6 * Copyright (c) 2008 Idiap Research Institute, http://www.idiap.ch/
7 * Written by Francois Fleuret <francois.fleuret@idiap.ch>
9 * This file is part of folded-ctf.
11 * folded-ctf is free software: you can redistribute it and/or modify
12 * it under the terms of the GNU General Public License version 3 as
13 * published by the Free Software Foundation.
15 * folded-ctf is distributed in the hope that it will be useful, but
16 * WITHOUT ANY WARRANTY; without even the implied warranty of
17 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
18 * General Public License for more details.
20 * You should have received a copy of the GNU General Public License
21 * along with folded-ctf. If not, see <http://www.gnu.org/licenses/>.
27 #include "fusion_sort.h"
29 scalar_t robust_sampling(int nb, scalar_t *weights, int nb_to_sample, int *sampled) {
32 for(int k = 0; k < nb_to_sample; k++) sampled[k] = 0;
35 scalar_t *pair_weights = new scalar_t[(nb+1)/2];
36 for(int k = 0; k < nb/2; k++)
37 pair_weights[k] = weights[2 * k] + weights[2 * k + 1];
39 pair_weights[(nb+1)/2 - 1] = weights[nb-1];
40 scalar_t result = robust_sampling((nb+1)/2, pair_weights, nb_to_sample, sampled);
41 for(int k = 0; k < nb_to_sample; k++) {
43 // There is a bit of a trick for the isolated sample in the odd
44 // case. Since the corresponding pair weight is the same as the
45 // one sample alone, the test is always true and the isolated
46 // sample will be taken for sure.
47 if(drand48() * pair_weights[s] <= weights[2 * s])
50 sampled[k] = 2 * s + 1;
52 delete[] pair_weights;
57 void print_roc_small_pos(ostream *out,
58 int nb_pos, scalar_t *pos_responses,
59 int nb_neg, scalar_t *neg_responses,
60 scalar_t fas_factor) {
62 scalar_t *sorted_pos_responses = new scalar_t[nb_pos];
64 fusion_sort(nb_pos, pos_responses, sorted_pos_responses);
66 int *bins = new int[nb_pos + 1];
67 for(int k = 0; k <= nb_pos; k++) bins[k] = 0;
69 for(int k = 0; k < nb_neg; k++) {
70 scalar_t r = neg_responses[k];
72 if(r < sorted_pos_responses[0])
75 else if(r >= sorted_pos_responses[nb_pos - 1])
85 if(r < sorted_pos_responses[c])
91 // Beware of identical positive responses
92 while(c < nb_pos && r >= sorted_pos_responses[c])
100 for(int k = 0; k < nb_pos; k++) {
102 if(k == 0 || sorted_pos_responses[k-1] < sorted_pos_responses[k]) {
103 (*out) << (scalar_t(s) / scalar_t(nb_neg)) * fas_factor
105 << scalar_t(nb_pos - k)/scalar_t(nb_pos)
107 << sorted_pos_responses[k]
113 delete[] sorted_pos_responses;