forked from ianbarber/php-svm
-
Notifications
You must be signed in to change notification settings - Fork 0
/
svm.c
1395 lines (1140 loc) · 37.6 KB
/
svm.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
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
/*
+----------------------------------------------------------------------+
| PHP Version 5 / svm |
+----------------------------------------------------------------------+
| Copyright (c) 2010 Ian Barber |
+----------------------------------------------------------------------+
| This source file is subject to version 3.0 of the PHP license, |
| that is bundled with this package in the file LICENSE, and is |
| available through the world-wide-web at the following url: |
| http://www.php.net/license/3_0.txt. |
| If you did not receive a copy of the PHP license and are unable to |
| obtain it through the world-wide-web, please send a note to |
| [email protected] so we can mail you a copy immediately. |
+----------------------------------------------------------------------+
| Authors: Ian Barber <[email protected]> |
+----------------------------------------------------------------------+
*/
#include "php_svm.h"
#include "php_svm_internal.h"
#include "php_ini.h" /* needed for 5.2 */
#include "Zend/zend_exceptions.h"
#include "ext/standard/info.h"
static zend_class_entry *php_svm_sc_entry;
static zend_class_entry *php_svm_model_sc_entry;
static zend_class_entry *php_svm_exception_sc_entry;
static zend_object_handlers svm_object_handlers;
static zend_object_handlers svm_model_object_handlers;
#ifndef TRUE
# define TRUE 1
# define FALSE 0
#endif
#define SVM_MAX_LINE_SIZE 4096
#define SVM_THROW(message, code) \
zend_throw_exception(php_svm_exception_sc_entry, message, (long)code TSRMLS_CC); \
return;
#define SVM_ERROR_MSG_SIZE 512
#define SVM_THROW_LAST_ERROR(fallback, code) \
zend_throw_exception(php_svm_exception_sc_entry, (strlen(intern->last_error) ? intern->last_error : fallback), (long)code TSRMLS_CC); \
memset(intern->last_error, 0, SVM_ERROR_MSG_SIZE); \
return;
typedef enum SvmLongAttribute {
SvmLongAttributeMin = 100,
phpsvm_svm_type,
phpsvm_kernel_type,
phpsvm_degree,
SvmLongAttributeMax /* Always add before this */
} SvmLongAttribute;
typedef enum SvmDoubleAttribute {
SvmDoubleAttributeMin = 200,
phpsvm_gamma,
phpsvm_nu,
phpsvm_eps,
phpsvm_p,
phpsvm_coef0,
phpsvm_C,
phpsvm_cache_size,
SvmDoubleAttributeMax /* Always add before this */
} SvmDoubleAttribute;
typedef enum SvmBoolAttribute {
SvmBoolAttributeMin = 300,
phpsvm_shrinking,
phpsvm_probability,
SvmBoolAttributeMax /* Always add before this */
} SvmBoolAttribute;
/* ---- START HELPER FUNCS ---- */
static void print_null(const char *s) {}
static zend_bool php_svm_set_bool_attribute(php_svm_object *intern, SvmBoolAttribute name, zend_bool value)
{
if (name >= SvmBoolAttributeMax) {
return FALSE;
}
switch (name) {
case phpsvm_shrinking:
intern->param.shrinking = value == TRUE ? 1 : 0;
break;
case phpsvm_probability:
intern->param.probability = value == TRUE ? 1 : 0;
break;
default:
return FALSE;
}
return TRUE;
}
static zend_bool php_svm_set_double_attribute(php_svm_object *intern, SvmDoubleAttribute name, double value)
{
if (name >= SvmDoubleAttributeMax) {
return FALSE;
}
switch (name) {
case phpsvm_gamma:
intern->param.gamma = value;
break;
case phpsvm_nu:
intern->param.nu = value;
break;
case phpsvm_eps:
intern->param.eps = value;
break;
case phpsvm_cache_size:
intern->param.cache_size = value;
break;
case phpsvm_p:
intern->param.p = value;
break;
case phpsvm_coef0:
intern->param.coef0 = value;
break;
case phpsvm_C:
intern->param.C = value;
break;
default:
return FALSE;
}
return TRUE;
}
static zend_bool php_svm_set_long_attribute(php_svm_object *intern, SvmLongAttribute name, long value)
{
if (name >= SvmLongAttributeMax) {
return FALSE;
}
switch (name) {
case phpsvm_svm_type:
if( value != C_SVC &&
value != NU_SVC &&
value != ONE_CLASS &&
value != EPSILON_SVR &&
value != NU_SVR ) {
return FALSE;
}
intern->param.svm_type = (int)value;
break;
case phpsvm_kernel_type:
if( value != LINEAR &&
value != POLY &&
value != RBF &&
value != SIGMOID &&
value != PRECOMPUTED ) {
return FALSE;
}
intern->param.kernel_type = (int)value;
break;
case phpsvm_degree:
intern->param.degree = (int)value;
break;
default:
return FALSE;
}
return TRUE;
}
/** {{{ zend_bool php_svm_stream_to_array(php_svm_object *intern, php_stream *stream, zval *retval TSRMLS_DC)
Take a stream containing lines of SVMLight format data and convert them into a PHP array for use by the training
function.
*/
static zend_bool php_svm_stream_to_array(php_svm_object *intern, php_stream *stream, zval *retval TSRMLS_DC)
{
while (!php_stream_eof(stream)) {
char buf[SVM_MAX_LINE_SIZE];
size_t retlen = 0;
int line = 1;
/* Read line by line */
if (php_stream_get_line(stream, buf, SVM_MAX_LINE_SIZE, &retlen)) {
zval *line_array, *pz_label;
char *label, *ptr, *l = NULL;
ptr = buf;
label = php_strtok_r(ptr, " \t", &l);
if (!label) {
snprintf(intern->last_error, SVM_ERROR_MSG_SIZE, "Incorrect data format on line %d", line);
return FALSE;
}
/* The line array */
MAKE_STD_ZVAL(line_array);
array_init(line_array);
/* The label */
MAKE_STD_ZVAL(pz_label);
ZVAL_STRING(pz_label, label, 1);
convert_to_double(pz_label);
/* Label is the first item in the line array */
add_next_index_zval(line_array, pz_label);
/* Read rest of the values on the line */
while (1) {
char *idx, *value;
zval *pz_idx, *pz_value;
/* idx:value format */
idx = php_strtok_r(NULL, ":", &l);
value = php_strtok_r(NULL, " \t", &l);
if (!value) {
break;
}
/* Make zvals and convert to correct types */
MAKE_STD_ZVAL(pz_idx);
ZVAL_STRING(pz_idx, idx, 1);
convert_to_long(pz_idx);
MAKE_STD_ZVAL(pz_value);
ZVAL_STRING(pz_value, value, 1);
convert_to_double(pz_value);
add_index_zval(line_array, Z_LVAL_P(pz_idx), pz_value);
zval_dtor(pz_idx);
FREE_ZVAL(pz_idx);
}
add_next_index_zval(retval, line_array);
line++;
}
}
return TRUE;
}
/* }}} */
/* {{{ int _php_count_values(zval *array);
For a an array of arrays, count the number of items in all subarrays.
*/
static int _php_count_values(zval *array)
{
zval **ppzval;
int values = 0;
for (zend_hash_internal_pointer_reset(Z_ARRVAL_P(array));
zend_hash_get_current_data(Z_ARRVAL_P(array), (void **) &ppzval) == SUCCESS;
zend_hash_move_forward(Z_ARRVAL_P(array))) {
if (Z_TYPE_PP(ppzval) == IS_ARRAY) {
values += zend_hash_num_elements(Z_ARRVAL_PP(ppzval));
}
}
return values;
}
/* }}} */
/* {{{ static void php_svm_free_problem(struct svm_problem *problem) {
Free the generated problem.
*/
static void php_svm_free_problem(struct svm_problem *problem) {
if (problem->x) {
efree(problem->x);
}
if (problem->y) {
efree(problem->y);
}
if (problem) {
efree(problem);
}
}
/* }}} */
/* {{{ static zend_bool php_svm_read_array(php_svm_object *intern, php_svm_model_object *intern_model, zval *array TSRMLS_DC)
Take a PHP array, and prepare libSVM problem data for training with.
*/
static struct svm_problem* php_svm_read_array(php_svm_object *intern, php_svm_model_object *intern_model, zval *array TSRMLS_DC)
{
zval **ppzval;
char *err_msg = NULL;
char *key;
char *endptr;
int i, num_labels, elements;
int j = 0, max_index = 0, inst_max_index = 0;
unsigned long index;
struct svm_problem *problem;
/* If reading multiple times make sure that we don't leak */
if (intern_model->x_space) {
efree(intern_model->x_space);
intern_model->x_space = NULL;
}
if (intern_model->model) {
#if LIBSVM_VERSION >= 300
svm_free_and_destroy_model(&intern_model->model);
#else
svm_destroy_model(intern_model->model);
#endif
intern_model->model = NULL;
}
/* Allocate the problem */
problem = emalloc(sizeof(struct svm_problem));
/* x and y */
num_labels = zend_hash_num_elements(HASH_OF(array));
/* Allocate space for the labels */
problem->y = emalloc(num_labels * sizeof(double));
/* allocate space for x */
problem->x = emalloc(num_labels * sizeof(struct svm_node *));
/* total number of elements */
elements = _php_count_values(array);
/* Allocate space: because the original array contains labels the alloc
should match as we add additional -1 node in the loop */
intern_model->x_space = emalloc(elements * sizeof(struct svm_node));
/* How many labels */
problem->l = num_labels;
/* Fill the problem */
for (zend_hash_internal_pointer_reset(Z_ARRVAL_P(array)), i = 0;
zend_hash_get_current_data(Z_ARRVAL_P(array), (void **) &ppzval) == SUCCESS;
zend_hash_move_forward(Z_ARRVAL_P(array)), i++) {
zval **ppz_label;
if (Z_TYPE_PP(ppzval) != IS_ARRAY) {
err_msg = "Data format error";
goto return_error;
}
if (zend_hash_num_elements(Z_ARRVAL_PP(ppzval)) < 2) {
err_msg = "Wrong amount of nodes in the sub-array";
goto return_error;
}
problem->x[i] = &(intern_model->x_space[j]);
zend_hash_internal_pointer_reset(Z_ARRVAL_PP(ppzval));
if ((zend_hash_get_current_data(Z_ARRVAL_PP(ppzval), (void **) &ppz_label) == SUCCESS)) {
if (Z_TYPE_PP(ppz_label) != IS_DOUBLE) {
convert_to_double(*ppz_label);
}
problem->y[i] = Z_DVAL_PP(ppz_label);
} else {
err_msg = "The sub-array contains only the label. Missing index-value pairs";
goto return_error;
}
while (1) {
zval **ppz_value;
if ((zend_hash_move_forward(Z_ARRVAL_PP(ppzval)) == SUCCESS) &&
(zend_hash_get_current_data(Z_ARRVAL_PP(ppzval), (void **) &ppz_value) == SUCCESS)) {
if (zend_hash_get_current_key(Z_ARRVAL_PP(ppzval), &key, &index, 0) == HASH_KEY_IS_STRING) {
intern_model->x_space[j].index = (int) strtol(key, &endptr, 10);
} else {
intern_model->x_space[j].index = (int) index;
}
if (Z_TYPE_PP(ppz_value) != IS_DOUBLE) {
convert_to_double(*ppz_value);
}
intern_model->x_space[j].value = Z_DVAL_PP(ppz_value);
inst_max_index = intern_model->x_space[j].index;
j++;
} else {
break;
}
}
intern_model->x_space[j++].index = -1;
if (inst_max_index > max_index) {
max_index = inst_max_index;
}
}
if (intern->param.gamma == 0 && max_index > 0) {
intern->param.gamma = 1.0/max_index;
}
return problem;
return_error:
php_svm_free_problem(problem);
if (err_msg) {
snprintf(intern->last_error, SVM_ERROR_MSG_SIZE, err_msg);
}
return NULL;
}
/* }}} */
/* {{{ static zend_bool php_svm_train(php_svm_object *intern, php_svm_model_object *intern_model, struct svm_problem *problem)
Train based on a libsvm problem structure
*/
static zend_bool php_svm_train(php_svm_object *intern, php_svm_model_object *intern_model, struct svm_problem *problem)
{
const char *err_msg = NULL;
err_msg = svm_check_parameter(problem, &(intern->param));
if (err_msg) {
snprintf(intern->last_error, SVM_ERROR_MSG_SIZE, err_msg);
return FALSE;
}
intern_model->model = svm_train(problem, &(intern->param));
/* Failure ? */
if (!intern_model->model) {
snprintf(intern->last_error, SVM_ERROR_MSG_SIZE, "Failed to train using the data");
return FALSE;
}
return TRUE;
}
/* }}} */
/* {{{ static zval* php_svm_get_data_from_param(php_svm_object *intern, zval *zparam)
Take an incoming parameter and convert it into a PHP array of svmlight style data.
*/
static zval* php_svm_get_data_from_param(php_svm_object *intern, zval *zparam TSRMLS_DC)
{
zval *data;
zend_bool our_stream = 0;
zend_bool need_read = 1;
php_stream *stream = NULL;
switch (Z_TYPE_P(zparam)) {
case IS_STRING:
stream = php_stream_open_wrapper(Z_STRVAL_P(zparam), "r", ENFORCE_SAFE_MODE | REPORT_ERRORS, NULL);
our_stream = 1;
break;
case IS_RESOURCE:
php_stream_from_zval_no_verify(stream, &zparam);
our_stream = 0;
break;
case IS_ARRAY:
our_stream = 0;
need_read = 0;
break;
default:
snprintf(intern->last_error, SVM_ERROR_MSG_SIZE, "Incorrect parameter type, expecting string, stream or an array");
return FALSE;
break;
}
/* If we got stream then read it in */
if (need_read) {
if (!stream) {
snprintf(intern->last_error, SVM_ERROR_MSG_SIZE, "Failed to open the data file");
return FALSE;
}
MAKE_STD_ZVAL(data);
array_init(data);
if (!php_svm_stream_to_array(intern, stream, data TSRMLS_CC)) {
zval_dtor(data);
FREE_ZVAL(data);
if (our_stream) {
php_stream_close(stream);
}
snprintf(intern->last_error, SVM_ERROR_MSG_SIZE, "Failed to read the data");
return FALSE;
}
} else {
data = zparam;
}
if (our_stream) {
php_stream_close(stream);
}
return data;
}
/* }}} */
/* {{{ static svm_node* php_svm_get_data_from_array(zval *arr)
Take an array of training data and turn it into an array of svm nodes.
*/
static struct svm_node* php_svm_get_data_from_array(zval* arr TSRMLS_DC)
{
struct svm_node *x;
zval **data;
HashTable *arr_hash;
HashPosition pointer;
int array_count, i;
char *endptr;
zval temp;
char *key;
uint key_len;
ulong index;
arr_hash = Z_ARRVAL_P(arr);
array_count = zend_hash_num_elements(arr_hash);
/* need 1 extra to indicate the end */
x = safe_emalloc((array_count + 1), sizeof(struct svm_node), 0);
i = 0;
/* Loop over the array in the argument and convert into svm_nodes for the prediction */
for(zend_hash_internal_pointer_reset_ex(arr_hash, &pointer);
zend_hash_get_current_data_ex(arr_hash, (void**) &data, &pointer) == SUCCESS;
zend_hash_move_forward_ex(arr_hash, &pointer))
{
if (zend_hash_get_current_key_ex(arr_hash, &key, &key_len, &index, 0, &pointer) == HASH_KEY_IS_STRING) {
x[i].index = (int) strtol(key, &endptr, 10);
} else {
x[i].index = (int) index;
}
temp = **data;
zval_copy_ctor(&temp);
convert_to_double(&temp);
x[i].value = Z_DVAL(temp);
zval_dtor(&temp);
i++;
}
/* needed so the predictor knows when to end */
x[i].index = -1;
return x;
}
/* }}} */
/* ---- END HELPER FUNCS ---- */
/* ---- START SVM ---- */
/* {{{ SVM SVM::__construct();
The constructor
*/
PHP_METHOD(svm, __construct)
{
php_svm_object *intern;
if (zend_parse_parameters(ZEND_NUM_ARGS() TSRMLS_CC, "") == FAILURE) {
SVM_THROW("Invalid parameters passed to constructor", 154);
}
intern = (php_svm_object *)zend_object_store_get_object(getThis() TSRMLS_CC);
/* Setup the default parameters to match those in libsvm's svm_train */
php_svm_set_long_attribute(intern, phpsvm_svm_type, C_SVC);
php_svm_set_long_attribute(intern, phpsvm_kernel_type, RBF);
php_svm_set_long_attribute(intern, phpsvm_degree, 3);
php_svm_set_double_attribute(intern, phpsvm_gamma, 0);
php_svm_set_double_attribute(intern, phpsvm_coef0, 0);
php_svm_set_double_attribute(intern, phpsvm_nu, 0.5);
php_svm_set_double_attribute(intern, phpsvm_cache_size, 100.0);
php_svm_set_double_attribute(intern, phpsvm_C, 1);
php_svm_set_double_attribute(intern, phpsvm_eps, 1e-3);
php_svm_set_double_attribute(intern, phpsvm_p, 0.1);
php_svm_set_bool_attribute(intern, phpsvm_shrinking, TRUE);
php_svm_set_bool_attribute(intern, phpsvm_probability, FALSE);
return;
}
/* }}} */
/* {{{ array SVM::getOptions();
Get training parameters, in an array.
*/
PHP_METHOD(svm, getOptions)
{
php_svm_object *intern;
intern = zend_object_store_get_object(getThis() TSRMLS_CC);
array_init(return_value);
add_index_long(return_value, phpsvm_svm_type, intern->param.svm_type);
add_index_long(return_value, phpsvm_kernel_type, intern->param.kernel_type);
add_index_long(return_value, phpsvm_degree, intern->param.degree);
add_index_long(return_value, phpsvm_coef0, intern->param.shrinking);
add_index_long(return_value, phpsvm_probability, intern->param.probability == 1 ? TRUE : FALSE);
add_index_long(return_value, phpsvm_shrinking, intern->param.shrinking == 1 ? TRUE : FALSE);
add_index_double(return_value, phpsvm_gamma, intern->param.gamma);
add_index_double(return_value, phpsvm_coef0, intern->param.coef0);
add_index_double(return_value, phpsvm_nu, intern->param.nu);
add_index_double(return_value, phpsvm_cache_size, intern->param.cache_size);
add_index_double(return_value, phpsvm_C, intern->param.C);
add_index_double(return_value, phpsvm_eps, intern->param.eps);
add_index_double(return_value, phpsvm_p, intern->param.p);
}
/* }}} */
/* {{{ int SVM::setOptopms(array params);
Takes an array of parameters and sets the training options to match them.
Only used by the training functions, will not modify an existing model.
*/
PHP_METHOD(svm, setOptions)
{
HashTable *params_ht;
php_svm_object *intern;
zval *params, **ppzval;
char *string_key = NULL;
ulong num_key;
if (zend_parse_parameters(ZEND_NUM_ARGS() TSRMLS_CC, "a", ¶ms) == FAILURE) {
RETURN_FALSE;
}
params_ht = HASH_OF(params);
if (zend_hash_num_elements(params_ht) == 0) {
return;
}
intern = (php_svm_object *)zend_object_store_get_object(getThis() TSRMLS_CC);
for (zend_hash_internal_pointer_reset(params_ht);
zend_hash_get_current_data(params_ht, (void **) &ppzval) == SUCCESS;
zend_hash_move_forward(params_ht)) {
zval tmp_zval, *tmp_pzval;
if (zend_hash_get_current_key(params_ht, &string_key, &num_key, 1) != HASH_KEY_IS_LONG) {
continue; /* Ignore the arg (TODO: throw exception?) */
}
/* Make sure we don't modify the original array */
tmp_zval = **ppzval;
zval_copy_ctor(&tmp_zval);
tmp_pzval = &tmp_zval;
/* Long attribute */
if (num_key > SvmLongAttributeMin && num_key < SvmLongAttributeMax) {
if (Z_TYPE_P(tmp_pzval) != IS_LONG) {
convert_to_long(tmp_pzval);
}
if (!php_svm_set_long_attribute(intern, num_key, Z_LVAL_P(tmp_pzval))) {
SVM_THROW("Failed to set the attribute", 999);
}
/* Double attribute */
} else if (num_key > SvmDoubleAttributeMin && num_key < SvmDoubleAttributeMax) {
if (Z_TYPE_P(tmp_pzval) != IS_DOUBLE) {
convert_to_double(tmp_pzval);
}
if (!php_svm_set_double_attribute(intern, num_key, Z_DVAL_P(tmp_pzval))) {
SVM_THROW("Failed to set the attribute", 999);
}
/* Bool attribute */
} else if(num_key > SvmBoolAttributeMin && num_key < SvmBoolAttributeMax) {
if (Z_TYPE_P(tmp_pzval) != IS_BOOL) {
convert_to_boolean(tmp_pzval);
}
if (!php_svm_set_bool_attribute(intern, num_key, Z_BVAL_P(tmp_pzval))) {
SVM_THROW("Failed to set the attribute", 999);
}
} else {
continue; /* Ignore the arg (TODO: throw exception?) */
}
tmp_pzval = NULL;
}
RETURN_TRUE;
}
/* }}} */
/* {{{ double SVM::crossvalidate(mixed string|resource|array, long folds);
Cross validate a the SVM parameters on the training data for tuning parameters. Will attempt to train then classify
on different segments of the training data (the total number of segments is the folds parameter). The training data
can be supplied as with the train function. For SVM classification, this will we return the correct percentage,
for regression the mean squared error.
@throws SVMException if the data format is incorrect
*/
PHP_METHOD(svm, crossvalidate)
{
int i;
int total_correct = 0;
long nrfolds;
double total_error = 0;
double sumv = 0, sumy = 0, sumvv = 0, sumyy = 0, sumvy = 0;
struct svm_problem *problem;
double returnval = 0.0;
double *target;
php_svm_object *intern;
php_svm_model_object *intern_return;
zval *zparam, *data, *tempobj;
if (zend_parse_parameters(ZEND_NUM_ARGS() TSRMLS_CC, "zl", &zparam, &nrfolds) == FAILURE) {
return;
}
intern = (php_svm_object *)zend_object_store_get_object(getThis() TSRMLS_CC);
ALLOC_INIT_ZVAL(tempobj);
object_init_ex(tempobj, php_svm_model_sc_entry);
intern_return = (php_svm_model_object *)zend_object_store_get_object(tempobj TSRMLS_CC);
data = php_svm_get_data_from_param(intern, zparam TSRMLS_CC);
if(!data) {
SVM_THROW_LAST_ERROR("Could not load data", 234);
}
problem = php_svm_read_array(intern, intern_return, data TSRMLS_CC);
if(!problem) {
SVM_THROW_LAST_ERROR("Cross validation failed", 1001);
}
target = emalloc(problem->l * sizeof(double));
svm_cross_validation(problem, &(intern->param), nrfolds, target);
if(intern->param.svm_type == EPSILON_SVR || intern->param.svm_type == NU_SVR) {
for(i=0;i<problem->l;i++) {
double y = problem->y[i];
double v = target[i];
total_error += (v-y)*(v-y);
sumv += v;
sumy += y;
sumvv += v*v;
sumyy += y*y;
sumvy += v*y;
}
returnval = (total_error/problem->l); // return total_error divded by number of examples
} else {
for(i=0; i<problem->l; i++) {
if(target[i] == problem->y[i]) {
++total_correct;
}
}
returnval = 1.0*total_correct/problem->l;
}
if (data != zparam) {
zval_dtor(data);
FREE_ZVAL(data);
}
zval_dtor(tempobj);
FREE_ZVAL(tempobj);
efree(target);
php_svm_free_problem(problem);
RETURN_DOUBLE(returnval);
}
/* }}} */
/* {{{ SVMModel SVM::train(mixed string|resource|array, [array classWeights]);
Train a SVM based on the SVMLight format data either in a file, an array, or in a previously opened stream.
@throws SVMException if the data format is incorrect. Can optionally accept a set of weights that will
be used to multiply C. Only useful for C_SVC kernels. These should be in the form array(class (int) => weight (float))
*/
PHP_METHOD(svm, train)
{
php_svm_object *intern;
php_svm_model_object *intern_return;
struct svm_problem *problem;
zval *data, *zparam, *weights, **ppzval;
HashTable *weights_ht;
int i;
char *key;
unsigned long index;
zend_bool status = 0;
weights = 0;
if (zend_parse_parameters(ZEND_NUM_ARGS() TSRMLS_CC, "z|a!", &zparam, &weights) == FAILURE) {
return;
}
intern = (php_svm_object *)zend_object_store_get_object(getThis() TSRMLS_CC);
if(weights && intern->param.svm_type != C_SVC) {
SVM_THROW("Weights can only be supplied for C_SVC training", 424);
}
data = php_svm_get_data_from_param(intern, zparam TSRMLS_CC);
if(!data) {
SVM_THROW_LAST_ERROR("Could not load data", 234);
}
if(weights) {
weights_ht = Z_ARRVAL_P(weights);
if(zend_hash_num_elements(weights_ht) > 0) {
intern->param.nr_weight = zend_hash_num_elements(weights_ht);
intern->param.weight_label = emalloc(intern->param.nr_weight * sizeof(int));
intern->param.weight = emalloc(intern->param.nr_weight * sizeof(double));
for (zend_hash_internal_pointer_reset(weights_ht), i = 0;
zend_hash_get_current_data(weights_ht, (void **) &ppzval) == SUCCESS;
zend_hash_move_forward(weights_ht), i++) {
zval tmp_zval, *tmp_pzval;
if (zend_hash_get_current_key(weights_ht, &key, &index, 0) == HASH_KEY_IS_LONG) {
intern->param.weight_label[i] = (int)index;
/* Make sure we don't modify the original array */
tmp_zval = **ppzval;
zval_copy_ctor(&tmp_zval);
tmp_pzval = &tmp_zval;
convert_to_double(tmp_pzval);
intern->param.weight[i] = Z_DVAL_P(tmp_pzval);
tmp_pzval = NULL;
}
}
}
} else {
intern->param.nr_weight = 0;
}
/* Return an object */
object_init_ex(return_value, php_svm_model_sc_entry);
intern_return = (php_svm_model_object *)zend_object_store_get_object(return_value TSRMLS_CC);
problem = php_svm_read_array(intern, intern_return, data TSRMLS_CC);
if(problem != NULL) {
if (php_svm_train(intern, intern_return, problem)) {
status = 1;
}
php_svm_free_problem(problem);
}
if(weights) {
efree(intern->param.weight_label);
efree(intern->param.weight);
}
if (data != zparam) {
zval_dtor(data);
FREE_ZVAL(data);
}
if (!status) {
SVM_THROW_LAST_ERROR("Training failed", 1000);
}
return;
}
/* }}} */
/* ---- END SVM ---- */
/* ---- START SVMMODEL ---- */
/** {{{ SvmModel::__construct([string filename])
Constructs an svm model
*/
PHP_METHOD(svmmodel, __construct)
{
php_svm_model_object *intern;
char *filename = NULL;
int filename_len;
if (zend_parse_parameters(ZEND_NUM_ARGS() TSRMLS_CC, "|s!", &filename, &filename_len) == FAILURE) {
SVM_THROW("Invalid parameters passed to constructor", 154);
}
if (!filename) {
return;
}
intern = (php_svm_model_object *)zend_object_store_get_object(getThis() TSRMLS_CC);
intern->model = svm_load_model(filename);
if (!intern->model) {
SVM_THROW("Failed to load the model", 1233);
}
return;
}
/* }}} */
/** {{{ SvmModel::load(string filename)
Loads the svm model from a file
*/
PHP_METHOD(svmmodel, load)
{
php_svm_model_object *intern;
char *filename = NULL;
int filename_len;
if (zend_parse_parameters(ZEND_NUM_ARGS() TSRMLS_CC, "s", &filename, &filename_len) == FAILURE) {
return;
}
intern = (php_svm_model_object *)zend_object_store_get_object(getThis() TSRMLS_CC);
intern->model = svm_load_model(filename);
if (!intern->model) {
SVM_THROW("Failed to load the model", 1233);
}
RETURN_TRUE;
}
/* }}} */
/** {{{ SvmModel::save(string filename)
Saves the svm model to a file
*/
PHP_METHOD(svmmodel, save)
{
php_svm_model_object *intern;
char *filename;
int filename_len;
if (zend_parse_parameters(ZEND_NUM_ARGS() TSRMLS_CC, "s", &filename, &filename_len) == FAILURE) {
return;
}
intern = (php_svm_model_object *)zend_object_store_get_object(getThis() TSRMLS_CC);
if (!intern->model) {
SVM_THROW("The object does not contain a model", 2321);
}
if (svm_save_model(filename, intern->model) != 0) {
SVM_THROW("Failed to save the model", 121);
}
RETURN_TRUE;
}
/* }}} */
/** {{{ SvmModel::getSvmType()
Gets the type of SVM the model was trained with
*/
PHP_METHOD(svmmodel, getSvmType)
{
php_svm_model_object *intern;
int svm_type;
intern = (php_svm_model_object *)zend_object_store_get_object(getThis() TSRMLS_CC);
if(!intern->model) {
SVM_THROW("No model available", 106);
}
svm_type = svm_get_svm_type(intern->model);
RETURN_LONG(svm_type);
}
/** {{{ SvmModel::getNrClass()
Gets the number of classes the model was trained with. Note that for a regression
or 1 class model 2 is returned.
*/
PHP_METHOD(svmmodel, getNrClass)
{
php_svm_model_object *intern;
int nr_classes;
intern = (php_svm_model_object *)zend_object_store_get_object(getThis() TSRMLS_CC);
if(!intern->model) {
SVM_THROW("No model available", 106);
}
nr_classes = svm_get_nr_class(intern->model);
RETURN_LONG(nr_classes);
}
/** {{{ SvmModel::getLabels()
Gets an array of labels that the model was trained with. For regression
and one class models, an empty array is returned.
*/
PHP_METHOD(svmmodel, getLabels)
{
php_svm_model_object *intern;
int nr_classes;
int* labels;
int i;
intern = (php_svm_model_object *)zend_object_store_get_object(getThis() TSRMLS_CC);
if(!intern->model) {
SVM_THROW("No model available", 106);
}
nr_classes = svm_get_nr_class(intern->model);
labels = safe_emalloc(nr_classes, sizeof(int), 0);
svm_get_labels(intern->model, labels);