ruby-fann 2.0.0 → 2.0.2
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +4 -4
- data/README.md +3 -2
- data/ext/ruby_fann/fann_augment.h +62 -62
- data/ext/ruby_fann/ruby_fann.c +362 -268
- data/lib/ruby_fann/version.rb +1 -1
- metadata +3 -3
data/ext/ruby_fann/ruby_fann.c
CHANGED
|
@@ -9,94 +9,131 @@ static VALUE m_rb_fann_standard_class;
|
|
|
9
9
|
static VALUE m_rb_fann_shortcut_class;
|
|
10
10
|
static VALUE m_rb_fann_train_data_class;
|
|
11
11
|
|
|
12
|
-
#define RETURN_FANN_INT(fn)
|
|
13
|
-
struct fann*
|
|
14
|
-
Data_Get_Struct
|
|
15
|
-
return INT2NUM(fn(f));
|
|
16
|
-
|
|
17
|
-
#define SET_FANN_INT(attr_name, fann_fn)
|
|
18
|
-
Check_Type(attr_name, T_FIXNUM);
|
|
19
|
-
struct fann*
|
|
20
|
-
Data_Get_Struct(self, struct fann, f); \
|
|
21
|
-
fann_fn(f, NUM2INT(attr_name));
|
|
22
|
-
return 0;
|
|
23
|
-
|
|
24
|
-
#define RETURN_FANN_UINT(fn)
|
|
25
|
-
struct fann*
|
|
26
|
-
Data_Get_Struct
|
|
27
|
-
return rb_int_new(fn(f));
|
|
28
|
-
|
|
29
|
-
#define SET_FANN_UINT(attr_name, fann_fn)
|
|
30
|
-
Check_Type(attr_name, T_FIXNUM);
|
|
31
|
-
struct fann*
|
|
32
|
-
Data_Get_Struct(self, struct fann, f); \
|
|
33
|
-
fann_fn(f, NUM2UINT(attr_name));
|
|
34
|
-
return 0;
|
|
12
|
+
#define RETURN_FANN_INT(fn) \
|
|
13
|
+
struct fann *f; \
|
|
14
|
+
Data_Get_Struct(self, struct fann, f); \
|
|
15
|
+
return INT2NUM(fn(f));
|
|
16
|
+
|
|
17
|
+
#define SET_FANN_INT(attr_name, fann_fn) \
|
|
18
|
+
Check_Type(attr_name, T_FIXNUM); \
|
|
19
|
+
struct fann *f; \
|
|
20
|
+
Data_Get_Struct(self, struct fann, f); \
|
|
21
|
+
fann_fn(f, NUM2INT(attr_name)); \
|
|
22
|
+
return 0;
|
|
23
|
+
|
|
24
|
+
#define RETURN_FANN_UINT(fn) \
|
|
25
|
+
struct fann *f; \
|
|
26
|
+
Data_Get_Struct(self, struct fann, f); \
|
|
27
|
+
return rb_int_new(fn(f));
|
|
28
|
+
|
|
29
|
+
#define SET_FANN_UINT(attr_name, fann_fn) \
|
|
30
|
+
Check_Type(attr_name, T_FIXNUM); \
|
|
31
|
+
struct fann *f; \
|
|
32
|
+
Data_Get_Struct(self, struct fann, f); \
|
|
33
|
+
fann_fn(f, NUM2UINT(attr_name)); \
|
|
34
|
+
return 0;
|
|
35
35
|
|
|
36
36
|
// Converts float return values to a double with same precision, avoids floating point errors.
|
|
37
|
-
#define RETURN_FANN_FLT(fn)
|
|
38
|
-
struct fann*
|
|
39
|
-
Data_Get_Struct
|
|
40
|
-
char buffy[20];
|
|
41
|
-
sprintf(buffy, "%0.6g", fn(f));
|
|
42
|
-
return rb_float_new(atof(buffy));
|
|
43
|
-
|
|
44
|
-
#define SET_FANN_FLT(attr_name, fann_fn)
|
|
45
|
-
Check_Type(attr_name, T_FLOAT);
|
|
46
|
-
struct fann*
|
|
47
|
-
Data_Get_Struct(self, struct fann, f); \
|
|
48
|
-
fann_fn(f, NUM2DBL(attr_name));
|
|
49
|
-
return self;
|
|
50
|
-
|
|
51
|
-
#define RETURN_FANN_DBL(fn)
|
|
52
|
-
struct fann*
|
|
53
|
-
Data_Get_Struct
|
|
54
|
-
return rb_float_new(fn(f));
|
|
37
|
+
#define RETURN_FANN_FLT(fn) \
|
|
38
|
+
struct fann *f; \
|
|
39
|
+
Data_Get_Struct(self, struct fann, f); \
|
|
40
|
+
char buffy[20]; \
|
|
41
|
+
sprintf(buffy, "%0.6g", fn(f)); \
|
|
42
|
+
return rb_float_new(atof(buffy));
|
|
43
|
+
|
|
44
|
+
#define SET_FANN_FLT(attr_name, fann_fn) \
|
|
45
|
+
Check_Type(attr_name, T_FLOAT); \
|
|
46
|
+
struct fann *f; \
|
|
47
|
+
Data_Get_Struct(self, struct fann, f); \
|
|
48
|
+
fann_fn(f, NUM2DBL(attr_name)); \
|
|
49
|
+
return self;
|
|
50
|
+
|
|
51
|
+
#define RETURN_FANN_DBL(fn) \
|
|
52
|
+
struct fann *f; \
|
|
53
|
+
Data_Get_Struct(self, struct fann, f); \
|
|
54
|
+
return rb_float_new(fn(f));
|
|
55
55
|
|
|
56
56
|
#define SET_FANN_DBL SET_FANN_FLT
|
|
57
57
|
|
|
58
58
|
// Convert ruby symbol to corresponding FANN enum type for activation function:
|
|
59
59
|
enum fann_activationfunc_enum sym_to_activation_function(VALUE activation_func)
|
|
60
60
|
{
|
|
61
|
-
ID id=SYM2ID(activation_func);
|
|
61
|
+
ID id = SYM2ID(activation_func);
|
|
62
62
|
enum fann_activationfunc_enum activation_function;
|
|
63
|
-
if(id==rb_intern("linear"))
|
|
63
|
+
if (id == rb_intern("linear"))
|
|
64
|
+
{
|
|
64
65
|
activation_function = FANN_LINEAR;
|
|
65
|
-
}
|
|
66
|
+
}
|
|
67
|
+
else if (id == rb_intern("threshold"))
|
|
68
|
+
{
|
|
66
69
|
activation_function = FANN_THRESHOLD;
|
|
67
|
-
}
|
|
70
|
+
}
|
|
71
|
+
else if (id == rb_intern("threshold_symmetric"))
|
|
72
|
+
{
|
|
68
73
|
activation_function = FANN_THRESHOLD_SYMMETRIC;
|
|
69
|
-
}
|
|
74
|
+
}
|
|
75
|
+
else if (id == rb_intern("sigmoid"))
|
|
76
|
+
{
|
|
70
77
|
activation_function = FANN_SIGMOID;
|
|
71
|
-
}
|
|
78
|
+
}
|
|
79
|
+
else if (id == rb_intern("sigmoid_stepwise"))
|
|
80
|
+
{
|
|
72
81
|
activation_function = FANN_SIGMOID_STEPWISE;
|
|
73
|
-
}
|
|
82
|
+
}
|
|
83
|
+
else if (id == rb_intern("sigmoid_symmetric"))
|
|
84
|
+
{
|
|
74
85
|
activation_function = FANN_SIGMOID_SYMMETRIC;
|
|
75
|
-
}
|
|
86
|
+
}
|
|
87
|
+
else if (id == rb_intern("sigmoid_symmetric_stepwise"))
|
|
88
|
+
{
|
|
76
89
|
activation_function = FANN_SIGMOID_SYMMETRIC_STEPWISE;
|
|
77
|
-
}
|
|
90
|
+
}
|
|
91
|
+
else if (id == rb_intern("gaussian"))
|
|
92
|
+
{
|
|
78
93
|
activation_function = FANN_GAUSSIAN;
|
|
79
|
-
}
|
|
94
|
+
}
|
|
95
|
+
else if (id == rb_intern("gaussian_symmetric"))
|
|
96
|
+
{
|
|
80
97
|
activation_function = FANN_GAUSSIAN_SYMMETRIC;
|
|
81
|
-
}
|
|
98
|
+
}
|
|
99
|
+
else if (id == rb_intern("gaussian_stepwise"))
|
|
100
|
+
{
|
|
82
101
|
activation_function = FANN_GAUSSIAN_STEPWISE;
|
|
83
|
-
}
|
|
102
|
+
}
|
|
103
|
+
else if (id == rb_intern("elliot"))
|
|
104
|
+
{
|
|
84
105
|
activation_function = FANN_ELLIOT;
|
|
85
|
-
}
|
|
106
|
+
}
|
|
107
|
+
else if (id == rb_intern("elliot_symmetric"))
|
|
108
|
+
{
|
|
86
109
|
activation_function = FANN_ELLIOT_SYMMETRIC;
|
|
87
|
-
}
|
|
110
|
+
}
|
|
111
|
+
else if (id == rb_intern("linear_piece"))
|
|
112
|
+
{
|
|
88
113
|
activation_function = FANN_LINEAR_PIECE;
|
|
89
|
-
}
|
|
114
|
+
}
|
|
115
|
+
else if (id == rb_intern("linear_piece_symmetric"))
|
|
116
|
+
{
|
|
90
117
|
activation_function = FANN_LINEAR_PIECE_SYMMETRIC;
|
|
91
|
-
}
|
|
118
|
+
}
|
|
119
|
+
else if (id == rb_intern("sin_symmetric"))
|
|
120
|
+
{
|
|
92
121
|
activation_function = FANN_SIN_SYMMETRIC;
|
|
93
|
-
}
|
|
122
|
+
}
|
|
123
|
+
else if (id == rb_intern("cos_symmetric"))
|
|
124
|
+
{
|
|
94
125
|
activation_function = FANN_COS_SYMMETRIC;
|
|
95
|
-
}
|
|
126
|
+
}
|
|
127
|
+
else if (id == rb_intern("sin"))
|
|
128
|
+
{
|
|
96
129
|
activation_function = FANN_SIN;
|
|
97
|
-
}
|
|
130
|
+
}
|
|
131
|
+
else if (id == rb_intern("cos"))
|
|
132
|
+
{
|
|
98
133
|
activation_function = FANN_COS;
|
|
99
|
-
}
|
|
134
|
+
}
|
|
135
|
+
else
|
|
136
|
+
{
|
|
100
137
|
rb_raise(rb_eRuntimeError, "Unrecognized activation function: [%s]", rb_id2name(SYM2ID(activation_func)));
|
|
101
138
|
}
|
|
102
139
|
return activation_function;
|
|
@@ -107,83 +144,118 @@ VALUE activation_function_to_sym(enum fann_activationfunc_enum fn)
|
|
|
107
144
|
{
|
|
108
145
|
VALUE activation_function;
|
|
109
146
|
|
|
110
|
-
if(fn==FANN_LINEAR)
|
|
147
|
+
if (fn == FANN_LINEAR)
|
|
148
|
+
{
|
|
111
149
|
activation_function = ID2SYM(rb_intern("linear"));
|
|
112
|
-
}
|
|
150
|
+
}
|
|
151
|
+
else if (fn == FANN_THRESHOLD)
|
|
152
|
+
{
|
|
113
153
|
activation_function = ID2SYM(rb_intern("threshold"));
|
|
114
|
-
}
|
|
154
|
+
}
|
|
155
|
+
else if (fn == FANN_THRESHOLD_SYMMETRIC)
|
|
156
|
+
{
|
|
115
157
|
activation_function = ID2SYM(rb_intern("threshold_symmetric"));
|
|
116
|
-
}
|
|
158
|
+
}
|
|
159
|
+
else if (fn == FANN_SIGMOID)
|
|
160
|
+
{
|
|
117
161
|
activation_function = ID2SYM(rb_intern("sigmoid"));
|
|
118
|
-
}
|
|
162
|
+
}
|
|
163
|
+
else if (fn == FANN_SIGMOID_STEPWISE)
|
|
164
|
+
{
|
|
119
165
|
activation_function = ID2SYM(rb_intern("sigmoid_stepwise"));
|
|
120
|
-
}
|
|
166
|
+
}
|
|
167
|
+
else if (fn == FANN_SIGMOID_SYMMETRIC)
|
|
168
|
+
{
|
|
121
169
|
activation_function = ID2SYM(rb_intern("sigmoid_symmetric"));
|
|
122
|
-
}
|
|
170
|
+
}
|
|
171
|
+
else if (fn == FANN_SIGMOID_SYMMETRIC_STEPWISE)
|
|
172
|
+
{
|
|
123
173
|
activation_function = ID2SYM(rb_intern("sigmoid_symmetric_stepwise"));
|
|
124
|
-
}
|
|
174
|
+
}
|
|
175
|
+
else if (fn == FANN_GAUSSIAN)
|
|
176
|
+
{
|
|
125
177
|
activation_function = ID2SYM(rb_intern("gaussian"));
|
|
126
|
-
}
|
|
178
|
+
}
|
|
179
|
+
else if (fn == FANN_GAUSSIAN_SYMMETRIC)
|
|
180
|
+
{
|
|
127
181
|
activation_function = ID2SYM(rb_intern("gaussian_symmetric"));
|
|
128
|
-
}
|
|
182
|
+
}
|
|
183
|
+
else if (fn == FANN_GAUSSIAN_STEPWISE)
|
|
184
|
+
{
|
|
129
185
|
activation_function = ID2SYM(rb_intern("gaussian_stepwise"));
|
|
130
|
-
}
|
|
186
|
+
}
|
|
187
|
+
else if (fn == FANN_ELLIOT)
|
|
188
|
+
{
|
|
131
189
|
activation_function = ID2SYM(rb_intern("elliot"));
|
|
132
|
-
}
|
|
190
|
+
}
|
|
191
|
+
else if (fn == FANN_ELLIOT_SYMMETRIC)
|
|
192
|
+
{
|
|
133
193
|
activation_function = ID2SYM(rb_intern("elliot_symmetric"));
|
|
134
|
-
}
|
|
194
|
+
}
|
|
195
|
+
else if (fn == FANN_LINEAR_PIECE)
|
|
196
|
+
{
|
|
135
197
|
activation_function = ID2SYM(rb_intern("linear_piece"));
|
|
136
|
-
}
|
|
198
|
+
}
|
|
199
|
+
else if (fn == FANN_LINEAR_PIECE_SYMMETRIC)
|
|
200
|
+
{
|
|
137
201
|
activation_function = ID2SYM(rb_intern("linear_piece_symmetric"));
|
|
138
|
-
}
|
|
202
|
+
}
|
|
203
|
+
else if (fn == FANN_SIN_SYMMETRIC)
|
|
204
|
+
{
|
|
139
205
|
activation_function = ID2SYM(rb_intern("sin_symmetric"));
|
|
140
|
-
}
|
|
206
|
+
}
|
|
207
|
+
else if (fn == FANN_COS_SYMMETRIC)
|
|
208
|
+
{
|
|
141
209
|
activation_function = ID2SYM(rb_intern("cos_symmetric"));
|
|
142
|
-
}
|
|
210
|
+
}
|
|
211
|
+
else if (fn == FANN_SIN)
|
|
212
|
+
{
|
|
143
213
|
activation_function = ID2SYM(rb_intern("sin"));
|
|
144
|
-
}
|
|
214
|
+
}
|
|
215
|
+
else if (fn == FANN_COS)
|
|
216
|
+
{
|
|
145
217
|
activation_function = ID2SYM(rb_intern("cos"));
|
|
146
|
-
}
|
|
218
|
+
}
|
|
219
|
+
else
|
|
220
|
+
{
|
|
147
221
|
rb_raise(rb_eRuntimeError, "Unrecognized activation function: [%d]", fn);
|
|
148
222
|
}
|
|
149
223
|
return activation_function;
|
|
150
224
|
}
|
|
151
225
|
|
|
152
|
-
|
|
153
226
|
// Unused for now:
|
|
154
|
-
static void fann_mark
|
|
227
|
+
static void fann_mark(struct fann *ann) {}
|
|
155
228
|
|
|
156
229
|
// #define DEBUG 1
|
|
157
230
|
|
|
158
231
|
// Free memory associated with FANN:
|
|
159
|
-
static void fann_free
|
|
232
|
+
static void fann_free(struct fann *ann)
|
|
160
233
|
{
|
|
161
|
-
|
|
234
|
+
fann_destroy(ann);
|
|
162
235
|
// ("Destroyed FANN network [%d].\n", ann);
|
|
163
236
|
}
|
|
164
237
|
|
|
165
238
|
// Free memory associated with FANN Training data:
|
|
166
|
-
static void fann_training_data_free
|
|
239
|
+
static void fann_training_data_free(struct fann_train_data *train_data)
|
|
167
240
|
{
|
|
168
|
-
|
|
241
|
+
fann_destroy_train(train_data);
|
|
169
242
|
// printf("Destroyed Training data [%d].\n", train_data);
|
|
170
243
|
}
|
|
171
244
|
|
|
172
245
|
// Create wrapper, but don't allocate anything...do that in
|
|
173
246
|
// initialize, so we can construct with args:
|
|
174
|
-
static VALUE fann_allocate
|
|
247
|
+
static VALUE fann_allocate(VALUE klass)
|
|
175
248
|
{
|
|
176
|
-
return Data_Wrap_Struct
|
|
249
|
+
return Data_Wrap_Struct(klass, fann_mark, fann_free, 0);
|
|
177
250
|
}
|
|
178
251
|
|
|
179
252
|
// Create wrapper, but don't allocate annything...do that in
|
|
180
253
|
// initialize, so we can construct with args:
|
|
181
|
-
static VALUE fann_training_data_allocate
|
|
254
|
+
static VALUE fann_training_data_allocate(VALUE klass)
|
|
182
255
|
{
|
|
183
|
-
return Data_Wrap_Struct
|
|
256
|
+
return Data_Wrap_Struct(klass, fann_mark, fann_training_data_free, 0);
|
|
184
257
|
}
|
|
185
258
|
|
|
186
|
-
|
|
187
259
|
// static VALUE invoke_training_callback(VALUE self)
|
|
188
260
|
// {
|
|
189
261
|
// VALUE callback = rb_funcall(self, rb_intern("training_callback"), 0);
|
|
@@ -194,8 +266,8 @@ static VALUE fann_training_data_allocate (VALUE klass)
|
|
|
194
266
|
// unsigned int max_epochs, unsigned int epochs_between_reports, float desired_error, unsigned int epochs)
|
|
195
267
|
|
|
196
268
|
static int FANN_API fann_training_callback(struct fann *ann, struct fann_train_data *train,
|
|
197
|
-
|
|
198
|
-
|
|
269
|
+
unsigned int max_epochs, unsigned int epochs_between_reports,
|
|
270
|
+
float desired_error, unsigned int epochs)
|
|
199
271
|
{
|
|
200
272
|
VALUE self = (VALUE)fann_get_user_data(ann);
|
|
201
273
|
VALUE args = rb_hash_new();
|
|
@@ -213,13 +285,13 @@ static int FANN_API fann_training_callback(struct fann *ann, struct fann_train_d
|
|
|
213
285
|
|
|
214
286
|
VALUE callback = rb_funcall(self, rb_intern("training_callback"), 1, args);
|
|
215
287
|
|
|
216
|
-
if (TYPE(callback)!=T_FIXNUM)
|
|
288
|
+
if (TYPE(callback) != T_FIXNUM)
|
|
217
289
|
{
|
|
218
|
-
rb_raise
|
|
290
|
+
rb_raise(rb_eRuntimeError, "Callback method must return an integer (-1 to stop training).");
|
|
219
291
|
}
|
|
220
292
|
|
|
221
293
|
int status = NUM2INT(callback);
|
|
222
|
-
if (status
|
|
294
|
+
if (status == -1)
|
|
223
295
|
{
|
|
224
296
|
printf("Callback method returned -1; training will stop.\n");
|
|
225
297
|
}
|
|
@@ -251,19 +323,19 @@ static VALUE fann_initialize(VALUE self, VALUE hash)
|
|
|
251
323
|
VALUE num_outputs = rb_hash_aref(hash, ID2SYM(rb_intern("num_outputs")));
|
|
252
324
|
VALUE hidden_neurons = rb_hash_aref(hash, ID2SYM(rb_intern("hidden_neurons")));
|
|
253
325
|
// printf("initializing\n\n\n");
|
|
254
|
-
struct fann*
|
|
255
|
-
if (TYPE(filename)==T_STRING)
|
|
326
|
+
struct fann *ann;
|
|
327
|
+
if (TYPE(filename) == T_STRING)
|
|
256
328
|
{
|
|
257
329
|
// Initialize with file:
|
|
258
330
|
// train_data = fann_read_train_from_file(StringValuePtr(filename));
|
|
259
331
|
// DATA_PTR(self) = train_data;
|
|
260
332
|
ann = fann_create_from_file(StringValuePtr(filename));
|
|
261
|
-
|
|
333
|
+
// printf("Created RubyFann::Standard [%d] from file [%s].\n", ann, StringValuePtr(filename));
|
|
262
334
|
}
|
|
263
|
-
else if(rb_obj_is_kind_of(self, m_rb_fann_shortcut_class))
|
|
335
|
+
else if (rb_obj_is_kind_of(self, m_rb_fann_shortcut_class))
|
|
264
336
|
{
|
|
265
337
|
// Initialize as shortcut, suitable for cascade training:
|
|
266
|
-
//ann = fann_create_shortcut_array(num_layers, layers);
|
|
338
|
+
// ann = fann_create_shortcut_array(num_layers, layers);
|
|
267
339
|
Check_Type(num_inputs, T_FIXNUM);
|
|
268
340
|
Check_Type(num_outputs, T_FIXNUM);
|
|
269
341
|
|
|
@@ -278,17 +350,18 @@ static VALUE fann_initialize(VALUE self, VALUE hash)
|
|
|
278
350
|
Check_Type(num_outputs, T_FIXNUM);
|
|
279
351
|
|
|
280
352
|
// Initialize layers:
|
|
281
|
-
unsigned int num_layers=RARRAY_LEN(hidden_neurons) + 2;
|
|
353
|
+
unsigned int num_layers = RARRAY_LEN(hidden_neurons) + 2;
|
|
282
354
|
unsigned int layers[num_layers];
|
|
283
355
|
|
|
284
356
|
// Input:
|
|
285
|
-
layers[0]=NUM2INT(num_inputs);
|
|
357
|
+
layers[0] = NUM2INT(num_inputs);
|
|
286
358
|
// Output:
|
|
287
|
-
layers[num_layers-1]=NUM2INT(num_outputs);
|
|
359
|
+
layers[num_layers - 1] = NUM2INT(num_outputs);
|
|
288
360
|
// Hidden:
|
|
289
361
|
unsigned int i;
|
|
290
|
-
for (i=1; i<=num_layers-2; i++)
|
|
291
|
-
|
|
362
|
+
for (i = 1; i <= num_layers - 2; i++)
|
|
363
|
+
{
|
|
364
|
+
layers[i] = NUM2INT(RARRAY_PTR(hidden_neurons)[i - 1]);
|
|
292
365
|
}
|
|
293
366
|
ann = fann_create_standard_array(num_layers, layers);
|
|
294
367
|
}
|
|
@@ -297,9 +370,9 @@ static VALUE fann_initialize(VALUE self, VALUE hash)
|
|
|
297
370
|
|
|
298
371
|
// printf("Checking for callback...");
|
|
299
372
|
|
|
300
|
-
//int callback = rb_protect(invoke_training_callback, (self), &status);
|
|
301
|
-
//
|
|
302
|
-
if(rb_respond_to(self, rb_intern("training_callback")))
|
|
373
|
+
// int callback = rb_protect(invoke_training_callback, (self), &status);
|
|
374
|
+
// VALUE callback = rb_funcall(DATA_PTR(self), "training_callback", 0);
|
|
375
|
+
if (rb_respond_to(self, rb_intern("training_callback")))
|
|
303
376
|
{
|
|
304
377
|
fann_set_callback(ann, &fann_training_callback);
|
|
305
378
|
fann_set_user_data(ann, self);
|
|
@@ -329,39 +402,39 @@ static VALUE fann_initialize(VALUE self, VALUE hash)
|
|
|
329
402
|
*/
|
|
330
403
|
static VALUE fann_train_data_initialize(VALUE self, VALUE hash)
|
|
331
404
|
{
|
|
332
|
-
struct fann_train_data*
|
|
405
|
+
struct fann_train_data *train_data;
|
|
333
406
|
Check_Type(hash, T_HASH);
|
|
334
407
|
|
|
335
408
|
VALUE filename = rb_hash_aref(hash, ID2SYM(rb_intern("filename")));
|
|
336
409
|
VALUE inputs = rb_hash_aref(hash, ID2SYM(rb_intern("inputs")));
|
|
337
410
|
VALUE desired_outputs = rb_hash_aref(hash, ID2SYM(rb_intern("desired_outputs")));
|
|
338
411
|
|
|
339
|
-
if (TYPE(filename)==T_STRING)
|
|
412
|
+
if (TYPE(filename) == T_STRING)
|
|
340
413
|
{
|
|
341
414
|
train_data = fann_read_train_from_file(StringValuePtr(filename));
|
|
342
415
|
DATA_PTR(self) = train_data;
|
|
343
416
|
}
|
|
344
|
-
else if (TYPE(inputs)==T_ARRAY)
|
|
417
|
+
else if (TYPE(inputs) == T_ARRAY)
|
|
345
418
|
{
|
|
346
|
-
if (TYPE(desired_outputs)!=T_ARRAY)
|
|
419
|
+
if (TYPE(desired_outputs) != T_ARRAY)
|
|
347
420
|
{
|
|
348
|
-
rb_raise
|
|
421
|
+
rb_raise(rb_eRuntimeError, "[desired_outputs] must be present when [inputs] used.");
|
|
349
422
|
}
|
|
350
423
|
|
|
351
424
|
if (RARRAY_LEN(inputs) < 1)
|
|
352
425
|
{
|
|
353
|
-
rb_raise
|
|
426
|
+
rb_raise(rb_eRuntimeError, "[inputs] must contain at least one value.");
|
|
354
427
|
}
|
|
355
428
|
|
|
356
429
|
if (RARRAY_LEN(desired_outputs) < 1)
|
|
357
430
|
{
|
|
358
|
-
rb_raise
|
|
431
|
+
rb_raise(rb_eRuntimeError, "[desired_outputs] must contain at least one value.");
|
|
359
432
|
}
|
|
360
433
|
|
|
361
434
|
// The data is here, start constructing:
|
|
362
|
-
if(RARRAY_LEN(inputs) != RARRAY_LEN(desired_outputs))
|
|
435
|
+
if (RARRAY_LEN(inputs) != RARRAY_LEN(desired_outputs))
|
|
363
436
|
{
|
|
364
|
-
rb_raise
|
|
437
|
+
rb_raise(
|
|
365
438
|
rb_eRuntimeError,
|
|
366
439
|
"Number of inputs must match number of outputs: (%d != %d)",
|
|
367
440
|
(int)RARRAY_LEN(inputs),
|
|
@@ -373,13 +446,12 @@ static VALUE fann_train_data_initialize(VALUE self, VALUE hash)
|
|
|
373
446
|
}
|
|
374
447
|
else
|
|
375
448
|
{
|
|
376
|
-
rb_raise
|
|
449
|
+
rb_raise(rb_eRuntimeError, "Must construct with a filename(string) or inputs/desired_outputs(arrays). All args passed via hash with symbols as keys.");
|
|
377
450
|
}
|
|
378
451
|
|
|
379
452
|
return (VALUE)train_data;
|
|
380
453
|
}
|
|
381
454
|
|
|
382
|
-
|
|
383
455
|
/** call-seq: save(filename)
|
|
384
456
|
|
|
385
457
|
Save to given filename
|
|
@@ -387,8 +459,8 @@ static VALUE fann_train_data_initialize(VALUE self, VALUE hash)
|
|
|
387
459
|
static VALUE training_save(VALUE self, VALUE filename)
|
|
388
460
|
{
|
|
389
461
|
Check_Type(filename, T_STRING);
|
|
390
|
-
struct fann_train_data*
|
|
391
|
-
Data_Get_Struct
|
|
462
|
+
struct fann_train_data *t;
|
|
463
|
+
Data_Get_Struct(self, struct fann_train_data, t);
|
|
392
464
|
fann_save_train(t, StringValuePtr(filename));
|
|
393
465
|
return self;
|
|
394
466
|
}
|
|
@@ -397,8 +469,8 @@ static VALUE training_save(VALUE self, VALUE filename)
|
|
|
397
469
|
This is recommended for incremental training, while it will have no influence during batch training.*/
|
|
398
470
|
static VALUE shuffle(VALUE self)
|
|
399
471
|
{
|
|
400
|
-
struct fann_train_data*
|
|
401
|
-
Data_Get_Struct
|
|
472
|
+
struct fann_train_data *t;
|
|
473
|
+
Data_Get_Struct(self, struct fann_train_data, t);
|
|
402
474
|
fann_shuffle_train_data(t);
|
|
403
475
|
return self;
|
|
404
476
|
}
|
|
@@ -406,9 +478,9 @@ static VALUE shuffle(VALUE self)
|
|
|
406
478
|
/** Length of training data*/
|
|
407
479
|
static VALUE length_train_data(VALUE self)
|
|
408
480
|
{
|
|
409
|
-
struct fann_train_data*
|
|
410
|
-
Data_Get_Struct
|
|
411
|
-
return(UINT2NUM(fann_length_train_data(t)));
|
|
481
|
+
struct fann_train_data *t;
|
|
482
|
+
Data_Get_Struct(self, struct fann_train_data, t);
|
|
483
|
+
return (UINT2NUM(fann_length_train_data(t)));
|
|
412
484
|
return self;
|
|
413
485
|
}
|
|
414
486
|
|
|
@@ -426,7 +498,7 @@ static VALUE set_activation_function(VALUE self, VALUE activation_func, VALUE la
|
|
|
426
498
|
Check_Type(layer, T_FIXNUM);
|
|
427
499
|
Check_Type(neuron, T_FIXNUM);
|
|
428
500
|
|
|
429
|
-
struct fann*
|
|
501
|
+
struct fann *f;
|
|
430
502
|
Data_Get_Struct(self, struct fann, f);
|
|
431
503
|
fann_set_activation_function(f, sym_to_activation_function(activation_func), NUM2INT(layer), NUM2INT(neuron));
|
|
432
504
|
return self;
|
|
@@ -442,7 +514,7 @@ static VALUE set_activation_function(VALUE self, VALUE activation_func, VALUE la
|
|
|
442
514
|
static VALUE set_activation_function_hidden(VALUE self, VALUE activation_func)
|
|
443
515
|
{
|
|
444
516
|
Check_Type(activation_func, T_SYMBOL);
|
|
445
|
-
struct fann*
|
|
517
|
+
struct fann *f;
|
|
446
518
|
Data_Get_Struct(self, struct fann, f);
|
|
447
519
|
fann_set_activation_function_hidden(f, sym_to_activation_function(activation_func));
|
|
448
520
|
return self;
|
|
@@ -463,7 +535,7 @@ static VALUE set_activation_function_layer(VALUE self, VALUE activation_func, VA
|
|
|
463
535
|
{
|
|
464
536
|
Check_Type(activation_func, T_SYMBOL);
|
|
465
537
|
Check_Type(layer, T_FIXNUM);
|
|
466
|
-
struct fann*
|
|
538
|
+
struct fann *f;
|
|
467
539
|
Data_Get_Struct(self, struct fann, f);
|
|
468
540
|
fann_set_activation_function_layer(f, sym_to_activation_function(activation_func), NUM2INT(layer));
|
|
469
541
|
return self;
|
|
@@ -480,7 +552,7 @@ static VALUE get_activation_function(VALUE self, VALUE layer, VALUE neuron)
|
|
|
480
552
|
{
|
|
481
553
|
Check_Type(layer, T_FIXNUM);
|
|
482
554
|
Check_Type(neuron, T_FIXNUM);
|
|
483
|
-
struct fann*
|
|
555
|
+
struct fann *f;
|
|
484
556
|
Data_Get_Struct(self, struct fann, f);
|
|
485
557
|
fann_type val = fann_get_activation_function(f, NUM2INT(layer), NUM2INT(neuron));
|
|
486
558
|
return activation_function_to_sym(val);
|
|
@@ -497,7 +569,7 @@ static VALUE get_activation_function(VALUE self, VALUE layer, VALUE neuron)
|
|
|
497
569
|
static VALUE set_activation_function_output(VALUE self, VALUE activation_func)
|
|
498
570
|
{
|
|
499
571
|
Check_Type(activation_func, T_SYMBOL);
|
|
500
|
-
struct fann*
|
|
572
|
+
struct fann *f;
|
|
501
573
|
Data_Get_Struct(self, struct fann, f);
|
|
502
574
|
fann_set_activation_function_output(f, sym_to_activation_function(activation_func));
|
|
503
575
|
return self;
|
|
@@ -511,7 +583,7 @@ static VALUE get_activation_steepness(VALUE self, VALUE layer, VALUE neuron)
|
|
|
511
583
|
{
|
|
512
584
|
Check_Type(layer, T_FIXNUM);
|
|
513
585
|
Check_Type(neuron, T_FIXNUM);
|
|
514
|
-
struct fann*
|
|
586
|
+
struct fann *f;
|
|
515
587
|
Data_Get_Struct(self, struct fann, f);
|
|
516
588
|
fann_type val = fann_get_activation_steepness(f, NUM2INT(layer), NUM2INT(neuron));
|
|
517
589
|
return rb_float_new(val);
|
|
@@ -527,7 +599,7 @@ static VALUE set_activation_steepness(VALUE self, VALUE steepness, VALUE layer,
|
|
|
527
599
|
Check_Type(layer, T_FIXNUM);
|
|
528
600
|
Check_Type(neuron, T_FIXNUM);
|
|
529
601
|
|
|
530
|
-
struct fann*
|
|
602
|
+
struct fann *f;
|
|
531
603
|
Data_Get_Struct(self, struct fann, f);
|
|
532
604
|
fann_set_activation_steepness(f, NUM2DBL(steepness), NUM2INT(layer), NUM2INT(neuron));
|
|
533
605
|
return self;
|
|
@@ -550,7 +622,7 @@ static VALUE set_activation_steepness_layer(VALUE self, VALUE steepness, VALUE l
|
|
|
550
622
|
Check_Type(steepness, T_FLOAT);
|
|
551
623
|
Check_Type(layer, T_FIXNUM);
|
|
552
624
|
|
|
553
|
-
struct fann*
|
|
625
|
+
struct fann *f;
|
|
554
626
|
Data_Get_Struct(self, struct fann, f);
|
|
555
627
|
fann_set_activation_steepness_layer(f, NUM2DBL(steepness), NUM2INT(layer));
|
|
556
628
|
return self;
|
|
@@ -684,9 +756,9 @@ static VALUE set_rprop_delta_zero(VALUE self, VALUE rprop_delta_zero)
|
|
|
684
756
|
/** Return array of bias(es)*/
|
|
685
757
|
static VALUE get_bias_array(VALUE self)
|
|
686
758
|
{
|
|
687
|
-
struct fann*
|
|
759
|
+
struct fann *f;
|
|
688
760
|
unsigned int num_layers;
|
|
689
|
-
Data_Get_Struct
|
|
761
|
+
Data_Get_Struct(self, struct fann, f);
|
|
690
762
|
num_layers = fann_get_num_layers(f);
|
|
691
763
|
unsigned int layers[num_layers];
|
|
692
764
|
fann_get_bias_array(f, layers);
|
|
@@ -695,7 +767,7 @@ static VALUE get_bias_array(VALUE self)
|
|
|
695
767
|
VALUE arr;
|
|
696
768
|
arr = rb_ary_new();
|
|
697
769
|
unsigned int i;
|
|
698
|
-
for (i=0; i<num_layers; i++)
|
|
770
|
+
for (i = 0; i < num_layers; i++)
|
|
699
771
|
{
|
|
700
772
|
rb_ary_push(arr, INT2NUM(layers[i]));
|
|
701
773
|
}
|
|
@@ -737,9 +809,9 @@ static VALUE get_neurons(VALUE self, VALUE layer)
|
|
|
737
809
|
struct fann_layer *layer_it;
|
|
738
810
|
struct fann_neuron *neuron_it;
|
|
739
811
|
|
|
740
|
-
struct fann*
|
|
812
|
+
struct fann *f;
|
|
741
813
|
unsigned int i;
|
|
742
|
-
Data_Get_Struct
|
|
814
|
+
Data_Get_Struct(self, struct fann, f);
|
|
743
815
|
|
|
744
816
|
VALUE neuron_array = rb_ary_new();
|
|
745
817
|
VALUE activation_function_sym = ID2SYM(rb_intern("activation_function"));
|
|
@@ -750,16 +822,17 @@ static VALUE get_neurons(VALUE self, VALUE layer)
|
|
|
750
822
|
VALUE connections_sym = ID2SYM(rb_intern("connections"));
|
|
751
823
|
unsigned int layer_num = 0;
|
|
752
824
|
|
|
753
|
-
|
|
754
|
-
|
|
755
|
-
for(layer_it = f->first_layer; layer_it != f->last_layer; layer_it++)
|
|
825
|
+
int nuke_bias_neuron = (fann_get_network_type(f) == FANN_NETTYPE_LAYER);
|
|
826
|
+
for (layer_it = f->first_layer; layer_it != f->last_layer; layer_it++)
|
|
756
827
|
{
|
|
757
|
-
for(neuron_it = layer_it->first_neuron; neuron_it != layer_it->last_neuron; neuron_it++)
|
|
828
|
+
for (neuron_it = layer_it->first_neuron; neuron_it != layer_it->last_neuron; neuron_it++)
|
|
758
829
|
{
|
|
759
|
-
if (nuke_bias_neuron && (neuron_it==(layer_it->last_neuron)-1))
|
|
830
|
+
if (nuke_bias_neuron && (neuron_it == (layer_it->last_neuron) - 1))
|
|
831
|
+
continue;
|
|
760
832
|
// Create array of connection indicies:
|
|
761
833
|
VALUE connection_array = rb_ary_new();
|
|
762
|
-
for (i = neuron_it->first_con; i < neuron_it->last_con; i++)
|
|
834
|
+
for (i = neuron_it->first_con; i < neuron_it->last_con; i++)
|
|
835
|
+
{
|
|
763
836
|
rb_ary_push(connection_array, INT2NUM(f->connections[i] - f->first_layer->first_neuron));
|
|
764
837
|
}
|
|
765
838
|
|
|
@@ -778,17 +851,16 @@ static VALUE get_neurons(VALUE self, VALUE layer)
|
|
|
778
851
|
++layer_num;
|
|
779
852
|
}
|
|
780
853
|
|
|
781
|
-
|
|
782
|
-
|
|
783
|
-
|
|
784
|
-
|
|
785
|
-
|
|
786
|
-
|
|
787
|
-
|
|
788
|
-
|
|
789
|
-
|
|
790
|
-
|
|
791
|
-
|
|
854
|
+
// switch (fann_get_network_type(ann)) {
|
|
855
|
+
// case FANN_NETTYPE_LAYER: {
|
|
856
|
+
// /* Report one bias in each layer except the last */
|
|
857
|
+
// if (layer_it != ann->last_layer-1)
|
|
858
|
+
// *bias = 1;
|
|
859
|
+
// else
|
|
860
|
+
// *bias = 0;
|
|
861
|
+
// break;
|
|
862
|
+
// }
|
|
863
|
+
// case FANN_NETTYPE_SHORTCUT: {
|
|
792
864
|
|
|
793
865
|
return neuron_array;
|
|
794
866
|
}
|
|
@@ -796,9 +868,9 @@ static VALUE get_neurons(VALUE self, VALUE layer)
|
|
|
796
868
|
/** Get list of layers in array format where each element contains number of neurons in that layer*/
|
|
797
869
|
static VALUE get_layer_array(VALUE self)
|
|
798
870
|
{
|
|
799
|
-
struct fann*
|
|
871
|
+
struct fann *f;
|
|
800
872
|
unsigned int num_layers;
|
|
801
|
-
Data_Get_Struct
|
|
873
|
+
Data_Get_Struct(self, struct fann, f);
|
|
802
874
|
num_layers = fann_get_num_layers(f);
|
|
803
875
|
unsigned int layers[num_layers];
|
|
804
876
|
fann_get_layer_array(f, layers);
|
|
@@ -807,12 +879,12 @@ static VALUE get_layer_array(VALUE self)
|
|
|
807
879
|
VALUE arr;
|
|
808
880
|
arr = rb_ary_new();
|
|
809
881
|
unsigned int i;
|
|
810
|
-
for (i=0; i<num_layers; i++)
|
|
882
|
+
for (i = 0; i < num_layers; i++)
|
|
811
883
|
{
|
|
812
884
|
rb_ary_push(arr, INT2NUM(layers[i]));
|
|
813
885
|
}
|
|
814
886
|
|
|
815
|
-
|
|
887
|
+
return arr;
|
|
816
888
|
}
|
|
817
889
|
|
|
818
890
|
/** Reads the mean square error from the network.*/
|
|
@@ -825,8 +897,8 @@ static VALUE get_MSE(VALUE self)
|
|
|
825
897
|
This function also resets the number of bits that fail.*/
|
|
826
898
|
static VALUE reset_MSE(VALUE self)
|
|
827
899
|
{
|
|
828
|
-
struct fann*
|
|
829
|
-
Data_Get_Struct
|
|
900
|
+
struct fann *f;
|
|
901
|
+
Data_Get_Struct(self, struct fann, f);
|
|
830
902
|
fann_reset_MSE(f);
|
|
831
903
|
return self;
|
|
832
904
|
}
|
|
@@ -834,18 +906,18 @@ static VALUE reset_MSE(VALUE self)
|
|
|
834
906
|
/** Get the type of network. Returns as ruby symbol (one of :shortcut, :layer)*/
|
|
835
907
|
static VALUE get_network_type(VALUE self)
|
|
836
908
|
{
|
|
837
|
-
struct fann*
|
|
909
|
+
struct fann *f;
|
|
838
910
|
enum fann_nettype_enum net_type;
|
|
839
911
|
VALUE ret_val;
|
|
840
|
-
Data_Get_Struct
|
|
912
|
+
Data_Get_Struct(self, struct fann, f);
|
|
841
913
|
|
|
842
914
|
net_type = fann_get_network_type(f);
|
|
843
915
|
|
|
844
|
-
if(net_type==FANN_NETTYPE_LAYER)
|
|
916
|
+
if (net_type == FANN_NETTYPE_LAYER)
|
|
845
917
|
{
|
|
846
918
|
ret_val = ID2SYM(rb_intern("layer")); // (rb_str_new2("FANN_NETTYPE_LAYER"));
|
|
847
919
|
}
|
|
848
|
-
else if(net_type==FANN_NETTYPE_SHORTCUT)
|
|
920
|
+
else if (net_type == FANN_NETTYPE_SHORTCUT)
|
|
849
921
|
{
|
|
850
922
|
ret_val = ID2SYM(rb_intern("shortcut")); // (rb_str_new2("FANN_NETTYPE_SHORTCUT"));
|
|
851
923
|
}
|
|
@@ -890,19 +962,24 @@ static VALUE set_train_error_function(VALUE self, VALUE train_error_function)
|
|
|
890
962
|
{
|
|
891
963
|
Check_Type(train_error_function, T_SYMBOL);
|
|
892
964
|
|
|
893
|
-
ID id=SYM2ID(train_error_function);
|
|
965
|
+
ID id = SYM2ID(train_error_function);
|
|
894
966
|
enum fann_errorfunc_enum fann_train_error_function;
|
|
895
967
|
|
|
896
|
-
if(id==rb_intern("linear"))
|
|
968
|
+
if (id == rb_intern("linear"))
|
|
969
|
+
{
|
|
897
970
|
fann_train_error_function = FANN_ERRORFUNC_LINEAR;
|
|
898
|
-
}
|
|
971
|
+
}
|
|
972
|
+
else if (id == rb_intern("tanh"))
|
|
973
|
+
{
|
|
899
974
|
fann_train_error_function = FANN_ERRORFUNC_TANH;
|
|
900
|
-
}
|
|
975
|
+
}
|
|
976
|
+
else
|
|
977
|
+
{
|
|
901
978
|
rb_raise(rb_eRuntimeError, "Unrecognized train error function: [%s]", rb_id2name(SYM2ID(train_error_function)));
|
|
902
979
|
}
|
|
903
980
|
|
|
904
|
-
struct fann*
|
|
905
|
-
Data_Get_Struct
|
|
981
|
+
struct fann *f;
|
|
982
|
+
Data_Get_Struct(self, struct fann, f);
|
|
906
983
|
fann_set_train_error_function(f, fann_train_error_function);
|
|
907
984
|
return self;
|
|
908
985
|
}
|
|
@@ -911,14 +988,14 @@ static VALUE set_train_error_function(VALUE self, VALUE train_error_function)
|
|
|
911
988
|
:linear, :tanh*/
|
|
912
989
|
static VALUE get_train_error_function(VALUE self)
|
|
913
990
|
{
|
|
914
|
-
struct fann*
|
|
991
|
+
struct fann *f;
|
|
915
992
|
enum fann_errorfunc_enum train_error;
|
|
916
993
|
VALUE ret_val;
|
|
917
|
-
Data_Get_Struct
|
|
994
|
+
Data_Get_Struct(self, struct fann, f);
|
|
918
995
|
|
|
919
996
|
train_error = fann_get_train_error_function(f);
|
|
920
997
|
|
|
921
|
-
if(train_error==FANN_ERRORFUNC_LINEAR)
|
|
998
|
+
if (train_error == FANN_ERRORFUNC_LINEAR)
|
|
922
999
|
{
|
|
923
1000
|
ret_val = ID2SYM(rb_intern("linear"));
|
|
924
1001
|
}
|
|
@@ -937,23 +1014,32 @@ static VALUE set_training_algorithm(VALUE self, VALUE train_error_function)
|
|
|
937
1014
|
{
|
|
938
1015
|
Check_Type(train_error_function, T_SYMBOL);
|
|
939
1016
|
|
|
940
|
-
ID id=SYM2ID(train_error_function);
|
|
1017
|
+
ID id = SYM2ID(train_error_function);
|
|
941
1018
|
enum fann_train_enum fann_train_algorithm;
|
|
942
1019
|
|
|
943
|
-
if(id==rb_intern("incremental"))
|
|
1020
|
+
if (id == rb_intern("incremental"))
|
|
1021
|
+
{
|
|
944
1022
|
fann_train_algorithm = FANN_TRAIN_INCREMENTAL;
|
|
945
|
-
}
|
|
1023
|
+
}
|
|
1024
|
+
else if (id == rb_intern("batch"))
|
|
1025
|
+
{
|
|
946
1026
|
fann_train_algorithm = FANN_TRAIN_BATCH;
|
|
947
|
-
}
|
|
1027
|
+
}
|
|
1028
|
+
else if (id == rb_intern("rprop"))
|
|
1029
|
+
{
|
|
948
1030
|
fann_train_algorithm = FANN_TRAIN_RPROP;
|
|
949
|
-
}
|
|
1031
|
+
}
|
|
1032
|
+
else if (id == rb_intern("quickprop"))
|
|
1033
|
+
{
|
|
950
1034
|
fann_train_algorithm = FANN_TRAIN_QUICKPROP;
|
|
951
|
-
}
|
|
1035
|
+
}
|
|
1036
|
+
else
|
|
1037
|
+
{
|
|
952
1038
|
rb_raise(rb_eRuntimeError, "Unrecognized training algorithm function: [%s]", rb_id2name(SYM2ID(train_error_function)));
|
|
953
1039
|
}
|
|
954
1040
|
|
|
955
|
-
struct fann*
|
|
956
|
-
Data_Get_Struct
|
|
1041
|
+
struct fann *f;
|
|
1042
|
+
Data_Get_Struct(self, struct fann, f);
|
|
957
1043
|
fann_set_training_algorithm(f, fann_train_algorithm);
|
|
958
1044
|
return self;
|
|
959
1045
|
}
|
|
@@ -962,20 +1048,27 @@ static VALUE set_training_algorithm(VALUE self, VALUE train_error_function)
|
|
|
962
1048
|
:incremental, :batch, :rprop, :quickprop */
|
|
963
1049
|
static VALUE get_training_algorithm(VALUE self)
|
|
964
1050
|
{
|
|
965
|
-
struct fann*
|
|
1051
|
+
struct fann *f;
|
|
966
1052
|
enum fann_train_enum fann_train_algorithm;
|
|
967
1053
|
VALUE ret_val;
|
|
968
|
-
Data_Get_Struct
|
|
1054
|
+
Data_Get_Struct(self, struct fann, f);
|
|
969
1055
|
|
|
970
1056
|
fann_train_algorithm = fann_get_training_algorithm(f);
|
|
971
1057
|
|
|
972
|
-
if(fann_train_algorithm==FANN_TRAIN_INCREMENTAL)
|
|
1058
|
+
if (fann_train_algorithm == FANN_TRAIN_INCREMENTAL)
|
|
1059
|
+
{
|
|
973
1060
|
ret_val = ID2SYM(rb_intern("incremental"));
|
|
974
|
-
}
|
|
1061
|
+
}
|
|
1062
|
+
else if (fann_train_algorithm == FANN_TRAIN_BATCH)
|
|
1063
|
+
{
|
|
975
1064
|
ret_val = ID2SYM(rb_intern("batch"));
|
|
976
|
-
}
|
|
1065
|
+
}
|
|
1066
|
+
else if (fann_train_algorithm == FANN_TRAIN_RPROP)
|
|
1067
|
+
{
|
|
977
1068
|
ret_val = ID2SYM(rb_intern("rprop"));
|
|
978
|
-
}
|
|
1069
|
+
}
|
|
1070
|
+
else if (fann_train_algorithm == FANN_TRAIN_QUICKPROP)
|
|
1071
|
+
{
|
|
979
1072
|
ret_val = ID2SYM(rb_intern("quickprop"));
|
|
980
1073
|
}
|
|
981
1074
|
return ret_val;
|
|
@@ -988,19 +1081,24 @@ static VALUE get_training_algorithm(VALUE self)
|
|
|
988
1081
|
static VALUE set_train_stop_function(VALUE self, VALUE train_stop_function)
|
|
989
1082
|
{
|
|
990
1083
|
Check_Type(train_stop_function, T_SYMBOL);
|
|
991
|
-
ID id=SYM2ID(train_stop_function);
|
|
1084
|
+
ID id = SYM2ID(train_stop_function);
|
|
992
1085
|
enum fann_stopfunc_enum fann_train_stop_function;
|
|
993
1086
|
|
|
994
|
-
if(id==rb_intern("mse"))
|
|
1087
|
+
if (id == rb_intern("mse"))
|
|
1088
|
+
{
|
|
995
1089
|
fann_train_stop_function = FANN_STOPFUNC_MSE;
|
|
996
|
-
}
|
|
1090
|
+
}
|
|
1091
|
+
else if (id == rb_intern("bit"))
|
|
1092
|
+
{
|
|
997
1093
|
fann_train_stop_function = FANN_STOPFUNC_BIT;
|
|
998
|
-
}
|
|
1094
|
+
}
|
|
1095
|
+
else
|
|
1096
|
+
{
|
|
999
1097
|
rb_raise(rb_eRuntimeError, "Unrecognized stop function: [%s]", rb_id2name(SYM2ID(train_stop_function)));
|
|
1000
1098
|
}
|
|
1001
1099
|
|
|
1002
|
-
struct fann*
|
|
1003
|
-
Data_Get_Struct
|
|
1100
|
+
struct fann *f;
|
|
1101
|
+
Data_Get_Struct(self, struct fann, f);
|
|
1004
1102
|
fann_set_train_stop_function(f, fann_train_stop_function);
|
|
1005
1103
|
return self;
|
|
1006
1104
|
}
|
|
@@ -1009,14 +1107,14 @@ static VALUE set_train_stop_function(VALUE self, VALUE train_stop_function)
|
|
|
1009
1107
|
:mse, :bit */
|
|
1010
1108
|
static VALUE get_train_stop_function(VALUE self)
|
|
1011
1109
|
{
|
|
1012
|
-
struct fann*
|
|
1110
|
+
struct fann *f;
|
|
1013
1111
|
enum fann_stopfunc_enum train_stop;
|
|
1014
1112
|
VALUE ret_val;
|
|
1015
|
-
Data_Get_Struct
|
|
1113
|
+
Data_Get_Struct(self, struct fann, f);
|
|
1016
1114
|
|
|
1017
1115
|
train_stop = fann_get_train_stop_function(f);
|
|
1018
1116
|
|
|
1019
|
-
if(train_stop==FANN_STOPFUNC_MSE)
|
|
1117
|
+
if (train_stop == FANN_STOPFUNC_MSE)
|
|
1020
1118
|
{
|
|
1021
1119
|
ret_val = ID2SYM(rb_intern("mse")); // (rb_str_new2("FANN_NETTYPE_LAYER"));
|
|
1022
1120
|
}
|
|
@@ -1027,13 +1125,12 @@ static VALUE get_train_stop_function(VALUE self)
|
|
|
1027
1125
|
return ret_val;
|
|
1028
1126
|
}
|
|
1029
1127
|
|
|
1030
|
-
|
|
1031
1128
|
/** Will print the connections of the ann in a compact matrix,
|
|
1032
1129
|
for easy viewing of the internals of the ann. */
|
|
1033
1130
|
static VALUE print_connections(VALUE self)
|
|
1034
1131
|
{
|
|
1035
|
-
struct fann*
|
|
1036
|
-
Data_Get_Struct
|
|
1132
|
+
struct fann *f;
|
|
1133
|
+
Data_Get_Struct(self, struct fann, f);
|
|
1037
1134
|
fann_print_connections(f);
|
|
1038
1135
|
return self;
|
|
1039
1136
|
}
|
|
@@ -1041,8 +1138,8 @@ static VALUE print_connections(VALUE self)
|
|
|
1041
1138
|
/** Print current NN parameters to stdout */
|
|
1042
1139
|
static VALUE print_parameters(VALUE self)
|
|
1043
1140
|
{
|
|
1044
|
-
struct fann*
|
|
1045
|
-
Data_Get_Struct
|
|
1141
|
+
struct fann *f;
|
|
1142
|
+
Data_Get_Struct(self, struct fann, f);
|
|
1046
1143
|
fann_print_parameters(f);
|
|
1047
1144
|
return Qnil;
|
|
1048
1145
|
}
|
|
@@ -1054,8 +1151,8 @@ static VALUE randomize_weights(VALUE self, VALUE min_weight, VALUE max_weight)
|
|
|
1054
1151
|
{
|
|
1055
1152
|
Check_Type(min_weight, T_FLOAT);
|
|
1056
1153
|
Check_Type(max_weight, T_FLOAT);
|
|
1057
|
-
struct fann*
|
|
1058
|
-
Data_Get_Struct
|
|
1154
|
+
struct fann *f;
|
|
1155
|
+
Data_Get_Struct(self, struct fann, f);
|
|
1059
1156
|
fann_randomize_weights(f, NUM2DBL(min_weight), NUM2DBL(max_weight));
|
|
1060
1157
|
return self;
|
|
1061
1158
|
}
|
|
@@ -1064,37 +1161,36 @@ static VALUE randomize_weights(VALUE self, VALUE min_weight, VALUE max_weight)
|
|
|
1064
1161
|
|
|
1065
1162
|
Run neural net on array<Float> of inputs with current parameters.
|
|
1066
1163
|
Returns array<Float> as output */
|
|
1067
|
-
static VALUE run
|
|
1164
|
+
static VALUE run(VALUE self, VALUE inputs)
|
|
1068
1165
|
{
|
|
1069
1166
|
Check_Type(inputs, T_ARRAY);
|
|
1070
1167
|
|
|
1071
|
-
|
|
1168
|
+
struct fann *f;
|
|
1072
1169
|
unsigned int i;
|
|
1073
|
-
fann_type*
|
|
1170
|
+
fann_type *outputs;
|
|
1074
1171
|
|
|
1075
1172
|
// Convert inputs to type needed for NN:
|
|
1076
1173
|
unsigned int len = RARRAY_LEN(inputs);
|
|
1077
1174
|
fann_type fann_inputs[len];
|
|
1078
|
-
for (i=0; i<len; i++)
|
|
1175
|
+
for (i = 0; i < len; i++)
|
|
1079
1176
|
{
|
|
1080
1177
|
fann_inputs[i] = NUM2DBL(RARRAY_PTR(inputs)[i]);
|
|
1081
1178
|
}
|
|
1082
1179
|
|
|
1083
|
-
|
|
1084
1180
|
// Obtain NN & run method:
|
|
1085
|
-
|
|
1181
|
+
Data_Get_Struct(self, struct fann, f);
|
|
1086
1182
|
outputs = fann_run(f, fann_inputs);
|
|
1087
1183
|
|
|
1088
1184
|
// Create ruby array & set outputs:
|
|
1089
1185
|
VALUE arr;
|
|
1090
1186
|
arr = rb_ary_new();
|
|
1091
|
-
unsigned int output_len=fann_get_num_output(f);
|
|
1092
|
-
for (i=0; i<output_len; i++)
|
|
1187
|
+
unsigned int output_len = fann_get_num_output(f);
|
|
1188
|
+
for (i = 0; i < output_len; i++)
|
|
1093
1189
|
{
|
|
1094
1190
|
rb_ary_push(arr, rb_float_new(outputs[i]));
|
|
1095
1191
|
}
|
|
1096
1192
|
|
|
1097
|
-
|
|
1193
|
+
return arr;
|
|
1098
1194
|
}
|
|
1099
1195
|
|
|
1100
1196
|
/** call-seq: init_weights(train_data) -> return value
|
|
@@ -1105,10 +1201,10 @@ static VALUE init_weights(VALUE self, VALUE train_data)
|
|
|
1105
1201
|
|
|
1106
1202
|
Check_Type(train_data, T_DATA);
|
|
1107
1203
|
|
|
1108
|
-
struct fann*
|
|
1109
|
-
struct fann_train_data*
|
|
1110
|
-
Data_Get_Struct
|
|
1111
|
-
Data_Get_Struct
|
|
1204
|
+
struct fann *f;
|
|
1205
|
+
struct fann_train_data *t;
|
|
1206
|
+
Data_Get_Struct(self, struct fann, f);
|
|
1207
|
+
Data_Get_Struct(train_data, struct fann_train_data, t);
|
|
1112
1208
|
|
|
1113
1209
|
fann_init_weights(f, t);
|
|
1114
1210
|
return self;
|
|
@@ -1124,7 +1220,7 @@ static VALUE train(VALUE self, VALUE input, VALUE expected_output)
|
|
|
1124
1220
|
Check_Type(input, T_ARRAY);
|
|
1125
1221
|
Check_Type(expected_output, T_ARRAY);
|
|
1126
1222
|
|
|
1127
|
-
struct fann*
|
|
1223
|
+
struct fann *f;
|
|
1128
1224
|
Data_Get_Struct(self, struct fann, f);
|
|
1129
1225
|
|
|
1130
1226
|
unsigned int num_input = RARRAY_LEN(input);
|
|
@@ -1134,11 +1230,13 @@ static VALUE train(VALUE self, VALUE input, VALUE expected_output)
|
|
|
1134
1230
|
|
|
1135
1231
|
unsigned int i;
|
|
1136
1232
|
|
|
1137
|
-
for (i = 0; i < num_input; i++)
|
|
1233
|
+
for (i = 0; i < num_input; i++)
|
|
1234
|
+
{
|
|
1138
1235
|
data_input[i] = NUM2DBL(RARRAY_PTR(input)[i]);
|
|
1139
1236
|
}
|
|
1140
1237
|
|
|
1141
|
-
for (i = 0; i < num_output; i++)
|
|
1238
|
+
for (i = 0; i < num_output; i++)
|
|
1239
|
+
{
|
|
1142
1240
|
data_output[i] = NUM2DBL(RARRAY_PTR(expected_output)[i]);
|
|
1143
1241
|
}
|
|
1144
1242
|
|
|
@@ -1161,10 +1259,10 @@ static VALUE train_on_data(VALUE self, VALUE train_data, VALUE max_epochs, VALUE
|
|
|
1161
1259
|
Check_Type(epochs_between_reports, T_FIXNUM);
|
|
1162
1260
|
Check_Type(desired_error, T_FLOAT);
|
|
1163
1261
|
|
|
1164
|
-
struct fann*
|
|
1165
|
-
struct fann_train_data*
|
|
1166
|
-
Data_Get_Struct
|
|
1167
|
-
Data_Get_Struct
|
|
1262
|
+
struct fann *f;
|
|
1263
|
+
struct fann_train_data *t;
|
|
1264
|
+
Data_Get_Struct(self, struct fann, f);
|
|
1265
|
+
Data_Get_Struct(train_data, struct fann_train_data, t);
|
|
1168
1266
|
|
|
1169
1267
|
unsigned int fann_max_epochs = NUM2INT(max_epochs);
|
|
1170
1268
|
unsigned int fann_epochs_between_reports = NUM2INT(epochs_between_reports);
|
|
@@ -1179,10 +1277,10 @@ static VALUE train_on_data(VALUE self, VALUE train_data, VALUE max_epochs, VALUE
|
|
|
1179
1277
|
static VALUE train_epoch(VALUE self, VALUE train_data)
|
|
1180
1278
|
{
|
|
1181
1279
|
Check_Type(train_data, T_DATA);
|
|
1182
|
-
struct fann*
|
|
1183
|
-
struct fann_train_data*
|
|
1184
|
-
Data_Get_Struct
|
|
1185
|
-
Data_Get_Struct
|
|
1280
|
+
struct fann *f;
|
|
1281
|
+
struct fann_train_data *t;
|
|
1282
|
+
Data_Get_Struct(self, struct fann, f);
|
|
1283
|
+
Data_Get_Struct(train_data, struct fann_train_data, t);
|
|
1186
1284
|
return rb_float_new(fann_train_epoch(f, t));
|
|
1187
1285
|
}
|
|
1188
1286
|
|
|
@@ -1192,10 +1290,10 @@ static VALUE train_epoch(VALUE self, VALUE train_data)
|
|
|
1192
1290
|
static VALUE test_data(VALUE self, VALUE train_data)
|
|
1193
1291
|
{
|
|
1194
1292
|
Check_Type(train_data, T_DATA);
|
|
1195
|
-
struct fann*
|
|
1196
|
-
struct fann_train_data*
|
|
1197
|
-
Data_Get_Struct
|
|
1198
|
-
Data_Get_Struct
|
|
1293
|
+
struct fann *f;
|
|
1294
|
+
struct fann_train_data *t;
|
|
1295
|
+
Data_Get_Struct(self, struct fann, f);
|
|
1296
|
+
Data_Get_Struct(train_data, struct fann_train_data, t);
|
|
1199
1297
|
return rb_float_new(fann_test_data(f, t));
|
|
1200
1298
|
}
|
|
1201
1299
|
|
|
@@ -1232,10 +1330,10 @@ static VALUE cascadetrain_on_data(VALUE self, VALUE train_data, VALUE max_neuron
|
|
|
1232
1330
|
Check_Type(neurons_between_reports, T_FIXNUM);
|
|
1233
1331
|
Check_Type(desired_error, T_FLOAT);
|
|
1234
1332
|
|
|
1235
|
-
struct fann*
|
|
1236
|
-
struct fann_train_data*
|
|
1237
|
-
Data_Get_Struct
|
|
1238
|
-
Data_Get_Struct
|
|
1333
|
+
struct fann *f;
|
|
1334
|
+
struct fann_train_data *t;
|
|
1335
|
+
Data_Get_Struct(self, struct fann, f);
|
|
1336
|
+
Data_Get_Struct(train_data, struct fann_train_data, t);
|
|
1239
1337
|
|
|
1240
1338
|
unsigned int fann_max_neurons = NUM2INT(max_neurons);
|
|
1241
1339
|
unsigned int fann_neurons_between_reports = NUM2INT(neurons_between_reports);
|
|
@@ -1305,7 +1403,6 @@ static VALUE set_cascade_candidate_stagnation_epochs(VALUE self, VALUE cascade_c
|
|
|
1305
1403
|
SET_FANN_UINT(cascade_candidate_stagnation_epochs, fann_set_cascade_candidate_stagnation_epochs);
|
|
1306
1404
|
}
|
|
1307
1405
|
|
|
1308
|
-
|
|
1309
1406
|
/** The weight multiplier is a parameter which is used to multiply the weights from the candidate neuron
|
|
1310
1407
|
before adding the neuron to the neural network. This parameter is usually between 0 and 1, and is used
|
|
1311
1408
|
to make the training a bit less aggressive. */
|
|
@@ -1426,13 +1523,13 @@ static VALUE set_learning_momentum(VALUE self, VALUE learning_momentum)
|
|
|
1426
1523
|
static VALUE set_cascade_activation_functions(VALUE self, VALUE cascade_activation_functions)
|
|
1427
1524
|
{
|
|
1428
1525
|
Check_Type(cascade_activation_functions, T_ARRAY);
|
|
1429
|
-
struct fann*
|
|
1430
|
-
Data_Get_Struct
|
|
1526
|
+
struct fann *f;
|
|
1527
|
+
Data_Get_Struct(self, struct fann, f);
|
|
1431
1528
|
|
|
1432
1529
|
unsigned long cnt = RARRAY_LEN(cascade_activation_functions);
|
|
1433
1530
|
enum fann_activationfunc_enum fann_activation_functions[cnt];
|
|
1434
1531
|
unsigned int i;
|
|
1435
|
-
for (i=0; i<cnt; i++)
|
|
1532
|
+
for (i = 0; i < cnt; i++)
|
|
1436
1533
|
{
|
|
1437
1534
|
fann_activation_functions[i] = sym_to_activation_function(RARRAY_PTR(cascade_activation_functions)[i]);
|
|
1438
1535
|
}
|
|
@@ -1445,16 +1542,16 @@ static VALUE set_cascade_activation_functions(VALUE self, VALUE cascade_activati
|
|
|
1445
1542
|
the candidates. The default is [:sigmoid, :sigmoid_symmetric, :gaussian, :gaussian_symmetric, :elliot, :elliot_symmetric] */
|
|
1446
1543
|
static VALUE get_cascade_activation_functions(VALUE self)
|
|
1447
1544
|
{
|
|
1448
|
-
struct fann*
|
|
1449
|
-
Data_Get_Struct
|
|
1545
|
+
struct fann *f;
|
|
1546
|
+
Data_Get_Struct(self, struct fann, f);
|
|
1450
1547
|
unsigned int cnt = fann_get_cascade_activation_functions_count(f);
|
|
1451
|
-
enum fann_activationfunc_enum*
|
|
1548
|
+
enum fann_activationfunc_enum *fann_functions = fann_get_cascade_activation_functions(f);
|
|
1452
1549
|
|
|
1453
1550
|
// Create ruby array & set outputs:
|
|
1454
1551
|
VALUE arr;
|
|
1455
1552
|
arr = rb_ary_new();
|
|
1456
1553
|
unsigned int i;
|
|
1457
|
-
for (i=0; i<cnt; i++)
|
|
1554
|
+
for (i = 0; i < cnt; i++)
|
|
1458
1555
|
{
|
|
1459
1556
|
rb_ary_push(arr, activation_function_to_sym(fann_functions[i]));
|
|
1460
1557
|
}
|
|
@@ -1490,13 +1587,13 @@ static VALUE set_cascade_num_candidate_groups(VALUE self, VALUE cascade_num_cand
|
|
|
1490
1587
|
static VALUE set_cascade_activation_steepnesses(VALUE self, VALUE cascade_activation_steepnesses)
|
|
1491
1588
|
{
|
|
1492
1589
|
Check_Type(cascade_activation_steepnesses, T_ARRAY);
|
|
1493
|
-
struct fann*
|
|
1494
|
-
Data_Get_Struct
|
|
1590
|
+
struct fann *f;
|
|
1591
|
+
Data_Get_Struct(self, struct fann, f);
|
|
1495
1592
|
|
|
1496
1593
|
unsigned int cnt = RARRAY_LEN(cascade_activation_steepnesses);
|
|
1497
1594
|
fann_type fann_activation_steepnesses[cnt];
|
|
1498
1595
|
unsigned int i;
|
|
1499
|
-
for (i=0; i<cnt; i++)
|
|
1596
|
+
for (i = 0; i < cnt; i++)
|
|
1500
1597
|
{
|
|
1501
1598
|
fann_activation_steepnesses[i] = NUM2DBL(RARRAY_PTR(cascade_activation_steepnesses)[i]);
|
|
1502
1599
|
}
|
|
@@ -1509,16 +1606,16 @@ static VALUE set_cascade_activation_steepnesses(VALUE self, VALUE cascade_activa
|
|
|
1509
1606
|
the candidates. */
|
|
1510
1607
|
static VALUE get_cascade_activation_steepnesses(VALUE self)
|
|
1511
1608
|
{
|
|
1512
|
-
struct fann*
|
|
1513
|
-
Data_Get_Struct
|
|
1514
|
-
fann_type*
|
|
1609
|
+
struct fann *f;
|
|
1610
|
+
Data_Get_Struct(self, struct fann, f);
|
|
1611
|
+
fann_type *fann_steepnesses = fann_get_cascade_activation_steepnesses(f);
|
|
1515
1612
|
unsigned int cnt = fann_get_cascade_activation_steepnesses_count(f);
|
|
1516
1613
|
|
|
1517
1614
|
// Create ruby array & set outputs:
|
|
1518
1615
|
VALUE arr;
|
|
1519
1616
|
arr = rb_ary_new();
|
|
1520
1617
|
unsigned int i;
|
|
1521
|
-
for (i=0; i<cnt; i++)
|
|
1618
|
+
for (i = 0; i < cnt; i++)
|
|
1522
1619
|
{
|
|
1523
1620
|
rb_ary_push(arr, rb_float_new(fann_steepnesses[i]));
|
|
1524
1621
|
}
|
|
@@ -1531,21 +1628,21 @@ static VALUE get_cascade_activation_steepnesses(VALUE self)
|
|
|
1531
1628
|
Save the entire network to configuration file with given name */
|
|
1532
1629
|
static VALUE nn_save(VALUE self, VALUE filename)
|
|
1533
1630
|
{
|
|
1534
|
-
struct fann*
|
|
1535
|
-
Data_Get_Struct
|
|
1631
|
+
struct fann *f;
|
|
1632
|
+
Data_Get_Struct(self, struct fann, f);
|
|
1536
1633
|
int status = fann_save(f, StringValuePtr(filename));
|
|
1537
1634
|
return INT2NUM(status);
|
|
1538
1635
|
}
|
|
1539
1636
|
|
|
1540
1637
|
/** Initializes class under RubyFann module/namespace. */
|
|
1541
|
-
void Init_ruby_fann
|
|
1638
|
+
void Init_ruby_fann()
|
|
1542
1639
|
{
|
|
1543
1640
|
// RubyFann module/namespace:
|
|
1544
|
-
m_rb_fann_module = rb_define_module
|
|
1641
|
+
m_rb_fann_module = rb_define_module("RubyFann");
|
|
1545
1642
|
|
|
1546
1643
|
// Standard NN class:
|
|
1547
|
-
m_rb_fann_standard_class = rb_define_class_under
|
|
1548
|
-
rb_define_alloc_func
|
|
1644
|
+
m_rb_fann_standard_class = rb_define_class_under(m_rb_fann_module, "Standard", rb_cObject);
|
|
1645
|
+
rb_define_alloc_func(m_rb_fann_standard_class, fann_allocate);
|
|
1549
1646
|
rb_define_method(m_rb_fann_standard_class, "initialize", fann_initialize, 1);
|
|
1550
1647
|
rb_define_method(m_rb_fann_standard_class, "init_weights", init_weights, 1);
|
|
1551
1648
|
rb_define_method(m_rb_fann_standard_class, "set_activation_function", set_activation_function, 3);
|
|
@@ -1608,7 +1705,6 @@ void Init_ruby_fann ()
|
|
|
1608
1705
|
rb_define_method(m_rb_fann_standard_class, "get_training_algorithm", get_training_algorithm, 0);
|
|
1609
1706
|
rb_define_method(m_rb_fann_standard_class, "set_training_algorithm", set_training_algorithm, 1);
|
|
1610
1707
|
|
|
1611
|
-
|
|
1612
1708
|
// Cascade functions:
|
|
1613
1709
|
rb_define_method(m_rb_fann_standard_class, "cascadetrain_on_data", cascadetrain_on_data, 4);
|
|
1614
1710
|
rb_define_method(m_rb_fann_standard_class, "get_cascade_output_change_fraction", get_cascade_output_change_fraction, 0);
|
|
@@ -1638,14 +1734,13 @@ void Init_ruby_fann ()
|
|
|
1638
1734
|
rb_define_method(m_rb_fann_standard_class, "set_cascade_num_candidate_groups", set_cascade_num_candidate_groups, 1);
|
|
1639
1735
|
rb_define_method(m_rb_fann_standard_class, "save", nn_save, 1);
|
|
1640
1736
|
|
|
1641
|
-
|
|
1642
1737
|
// Uncomment for fixed-point mode (also recompile fann). Probably not going to be needed:
|
|
1643
|
-
//rb_define_method(clazz, "get_decimal_point", get_decimal_point, 0);
|
|
1644
|
-
//rb_define_method(clazz, "get_multiplier", get_multiplier, 0);
|
|
1738
|
+
// rb_define_method(clazz, "get_decimal_point", get_decimal_point, 0);
|
|
1739
|
+
// rb_define_method(clazz, "get_multiplier", get_multiplier, 0);
|
|
1645
1740
|
|
|
1646
1741
|
// Shortcut NN class (duplicated from above so that rdoc generation tools can find the methods:):
|
|
1647
|
-
m_rb_fann_shortcut_class = rb_define_class_under
|
|
1648
|
-
rb_define_alloc_func
|
|
1742
|
+
m_rb_fann_shortcut_class = rb_define_class_under(m_rb_fann_module, "Shortcut", rb_cObject);
|
|
1743
|
+
rb_define_alloc_func(m_rb_fann_shortcut_class, fann_allocate);
|
|
1649
1744
|
rb_define_method(m_rb_fann_shortcut_class, "initialize", fann_initialize, 1);
|
|
1650
1745
|
rb_define_method(m_rb_fann_shortcut_class, "init_weights", init_weights, 1);
|
|
1651
1746
|
rb_define_method(m_rb_fann_shortcut_class, "set_activation_function", set_activation_function, 3);
|
|
@@ -1737,10 +1832,9 @@ void Init_ruby_fann ()
|
|
|
1737
1832
|
rb_define_method(m_rb_fann_shortcut_class, "set_cascade_num_candidate_groups", set_cascade_num_candidate_groups, 1);
|
|
1738
1833
|
rb_define_method(m_rb_fann_shortcut_class, "save", nn_save, 1);
|
|
1739
1834
|
|
|
1740
|
-
|
|
1741
1835
|
// TrainData NN class:
|
|
1742
|
-
m_rb_fann_train_data_class = rb_define_class_under
|
|
1743
|
-
rb_define_alloc_func
|
|
1836
|
+
m_rb_fann_train_data_class = rb_define_class_under(m_rb_fann_module, "TrainData", rb_cObject);
|
|
1837
|
+
rb_define_alloc_func(m_rb_fann_train_data_class, fann_training_data_allocate);
|
|
1744
1838
|
rb_define_method(m_rb_fann_train_data_class, "initialize", fann_train_data_initialize, 1);
|
|
1745
1839
|
rb_define_method(m_rb_fann_train_data_class, "length", length_train_data, 0);
|
|
1746
1840
|
rb_define_method(m_rb_fann_train_data_class, "shuffle", shuffle, 0);
|