-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathknntest.c
215 lines (176 loc) · 6.06 KB
/
knntest.c
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
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
/************************************************************************
* *
* Program package 'lvq_pak': *
* *
* knntest.c *
* -displays the recognition accuracy by knn test *
* *
* Version 3.0 *
* Date: 1 Mar 1995 *
* *
* NOTE: This program package is copyrighted in the sense that it *
* may be used for scientific purposes. The package as a whole, or *
* parts thereof, cannot be included or used in any commercial *
* application without written permission granted by its producents. *
* No programs contained in this package may be copied for commercial *
* distribution. *
* *
* All comments concerning this program package may be sent to the *
* e-mail address 'lvq@cochlea.hut.fi'. *
* *
************************************************************************/
#include <stdio.h>
#include <float.h>
#include "lvq_pak.h"
#include "errors.h"
#include "labels.h"
#include "datafile.h"
static char *usage[] = {
"knntest - displays the recognition accuracy by knn test\n",
"Required parameters:\n",
" -cin filename input codebook file\n",
" -din filename test data\n",
"Optional parameters:\n",
" -knn N use N nearest neighbours (default: 5)\n",
" -buffer integer buffered reading of data, integer lines at a time\n",
" -selfuncs name select a set of functions\n",
NULL};
int compute_knnaccuracy(struct teach_params *teach)
{
long noc, i, total, stotal;
struct data_entry *datatmp;
struct winner_info *winners;
WINNER_FUNCTION *find_winners = teach->winner;
int datalabel;
struct entries *data = teach->data;
struct entries *codes = teach->codes;
int knn = teach->knn;
struct hitlist *hits, *correct, *totals;
struct hit_entry *he;
eptr p;
if (knn < 1)
knn = 1;
winners = calloc(knn, sizeof(struct winner_info));
if (winners == NULL)
return ERR_NOMEM;
if ((hits = new_hitlist()) == NULL)
{
free(winners);
return ERR_NOMEM;
}
if ((correct = new_hitlist()) == NULL)
{
free(winners);
free_hitlist(hits);
return ERR_NOMEM;
}
if ((totals = new_hitlist()) == NULL)
{
free(winners);
free_hitlist(hits);
free_hitlist(correct);
return ERR_NOMEM;
}
stotal = 0;
total = 0;
ifverbose(3)
fprintf(stderr, "computing accuracy\n");
/* Scan all input entries */
datatmp = rewind_entries(data, &p);
/* Number of data vectors */
noc = data->flags.totlen_known ? data->num_entries : 0;
while (datatmp != NULL) {
find_winners(codes, datatmp, winners, knn);
/* If classification was correct */
clear_hitlist(hits);
for (i = 0; i < knn; i++)
add_hit(hits, get_entry_label(winners[i].winner));
datalabel = get_entry_label(datatmp);
if (hits->head->label == datalabel)
{
/* Number of correct classifications */
stotal++;
/* Number of correct classifications in that class */
add_hit(correct, datalabel);
}
/* Total number of entries in that class */
add_hit(totals, datalabel);
/* Total number of entries */
total++;
/* Take the next input entry */
datatmp = next_entry(&p);
ifverbose(1)
if (noc)
mprint(noc--);
}
ifverbose(1)
{
mprint((long) 0);
fprintf(stderr, "\n");
}
fprintf(stdout, "\nRecognition accuracy:\n\n");
for (he = totals->head; he != NULL; he = he->next)
{
int res, tot;
tot = he->freq;
res = hitlist_label_freq(correct, he->label);
fprintf(stdout, "%14s: ", find_conv_to_lab(he->label));
fprintf(stdout, "%6.2f %%\n", 100.0 * (float) res / tot);
}
fprintf(stdout, "\nTotal accuracy: %6.2f %%\n\n",
100.0 * (float) stotal / total);
free(winners);
free_hitlist(hits);
free_hitlist(correct);
free_hitlist(totals);
return 0;
}
int main(int argc, char **argv)
{
int knn;
long buffer;
char *in_data_file;
char *in_code_file;
struct entries *data, *codes;
struct teach_params params;
char *funcname = NULL;
global_options(argc, argv);
if (extract_parameter(argc, argv, "-help", OPTION2))
{
printhelp();
exit(0);
}
in_data_file = extract_parameter(argc, argv, IN_DATA_FILE, ALWAYS);
in_code_file = extract_parameter(argc, argv, IN_CODE_FILE, ALWAYS);
knn = (int) oatoi(extract_parameter(argc, argv, KNN_NEIGHBORS, OPTION), 5);
buffer = oatoi(extract_parameter(argc, argv, "-buffer", OPTION), 0);
funcname = extract_parameter(argc, argv, "-selfuncs", OPTION);
ifverbose(2)
fprintf(stderr, "Input entries are read from file %s\n", in_data_file);
if ((data = open_entries(in_data_file)) == NULL)
{
fprintf(stderr, "Can't open data file %s\n", in_data_file);
exit(1);
}
ifverbose(2)
fprintf(stderr, "Codebook entries are read from file %s\n", in_code_file);
if ((codes = open_entries(in_code_file)) == NULL)
{
fprintf(stderr, "Can't open codes file %s\n", in_code_file);
close_entries(data);
exit(1);
}
if (data->dimension != codes->dimension) {
fprintf(stderr, "Data and codebook vectors have different dimensions");
close_entries(data);
close_entries(codes);
exit(1);
}
set_teach_params(¶ms, codes, data, buffer, funcname);
params.winner = find_winner_knn;
params.knn = knn;
compute_knnaccuracy(¶ms);
close_entries(data);
close_entries(codes);
return(0);
}