@@ -3297,38 +3297,6 @@ fail: SWIGINTERN VALUE -_wrap_svm_get_obj(int argc, VALUE *argv, VALUE self) { - svm_model *arg1 = (svm_model *) 0 ; - int arg2 ; - void *argp1 = 0 ; - int res1 = 0 ; - int val2 ; - int ecode2 = 0 ; - double result; - VALUE vresult = Qnil; - - if ((argc < 2) || (argc > 2)) { - rb_raise(rb_eArgError, "wrong # of arguments(%d for 2)",argc); SWIG_fail; - } - res1 = SWIG_ConvertPtr(argv[0], &argp1,SWIGTYPE_p_svm_model, 0 | 0 ); - if (!SWIG_IsOK(res1)) { - SWIG_exception_fail(SWIG_ArgError(res1), Ruby_Format_TypeError( "", "svm_model const *","svm_get_obj", 1, argv[0] )); - } - arg1 = reinterpret_cast< svm_model * >(argp1); - ecode2 = SWIG_AsVal_int(argv[1], &val2); - if (!SWIG_IsOK(ecode2)) { - SWIG_exception_fail(SWIG_ArgError(ecode2), Ruby_Format_TypeError( "", "int","svm_get_obj", 2, argv[1] )); - } - arg2 = static_cast< int >(val2); - result = (double)svm_get_obj((svm_model const *)arg1,arg2); - vresult = SWIG_From_double(static_cast< double >(result)); - return vresult; -fail: - return Qnil; -} - - -SWIGINTERN VALUE _wrap_svm_get_svm_type(int argc, VALUE *argv, VALUE self) { svm_model *arg1 = (svm_model *) 0 ; void *argp1 = 0 ; @@ -3616,112 +3584,6 @@ fail: SWIGINTERN VALUE -_wrap_svm_get_model_rho(int argc, VALUE *argv, VALUE self) { - svm_model *arg1 = (svm_model *) 0 ; - void *argp1 = 0 ; - int res1 = 0 ; - double result; - VALUE vresult = Qnil; - - if ((argc < 1) || (argc > 1)) { - rb_raise(rb_eArgError, "wrong # of arguments(%d for 1)",argc); SWIG_fail; - } - res1 = SWIG_ConvertPtr(argv[0], &argp1,SWIGTYPE_p_svm_model, 0 | 0 ); - if (!SWIG_IsOK(res1)) { - SWIG_exception_fail(SWIG_ArgError(res1), Ruby_Format_TypeError( "", "svm_model *","svm_get_model_rho", 1, argv[0] )); - } - arg1 = reinterpret_cast< svm_model * >(argp1); - result = (double)svm_get_model_rho(arg1); - vresult = SWIG_From_double(static_cast< double >(result)); - return vresult; -fail: - return Qnil; -} - - -SWIGINTERN VALUE -_wrap_svm_get_model_num_coefs(int argc, VALUE *argv, VALUE self) { - svm_model *arg1 = (svm_model *) 0 ; - void *argp1 = 0 ; - int res1 = 0 ; - int result; - VALUE vresult = Qnil; - - if ((argc < 1) || (argc > 1)) { - rb_raise(rb_eArgError, "wrong # of arguments(%d for 1)",argc); SWIG_fail; - } - res1 = SWIG_ConvertPtr(argv[0], &argp1,SWIGTYPE_p_svm_model, 0 | 0 ); - if (!SWIG_IsOK(res1)) { - SWIG_exception_fail(SWIG_ArgError(res1), Ruby_Format_TypeError( "", "svm_model *","svm_get_model_num_coefs", 1, argv[0] )); - } - arg1 = reinterpret_cast< svm_model * >(argp1); - result = (int)svm_get_model_num_coefs(arg1); - vresult = SWIG_From_int(static_cast< int >(result)); - return vresult; -fail: - return Qnil; -} - - -SWIGINTERN VALUE -_wrap_svm_get_model_coefs(int argc, VALUE *argv, VALUE self) { - svm_model *arg1 = (svm_model *) 0 ; - double *arg2 = (double *) 0 ; - void *argp1 = 0 ; - int res1 = 0 ; - void *argp2 = 0 ; - int res2 = 0 ; - - if ((argc < 2) || (argc > 2)) { - rb_raise(rb_eArgError, "wrong # of arguments(%d for 2)",argc); SWIG_fail; - } - res1 = SWIG_ConvertPtr(argv[0], &argp1,SWIGTYPE_p_svm_model, 0 | 0 ); - if (!SWIG_IsOK(res1)) { - SWIG_exception_fail(SWIG_ArgError(res1), Ruby_Format_TypeError( "", "svm_model *","svm_get_model_coefs", 1, argv[0] )); - } - arg1 = reinterpret_cast< svm_model * >(argp1); - res2 = SWIG_ConvertPtr(argv[1], &argp2,SWIGTYPE_p_double, 0 | 0 ); - if (!SWIG_IsOK(res2)) { - SWIG_exception_fail(SWIG_ArgError(res2), Ruby_Format_TypeError( "", "double *","svm_get_model_coefs", 2, argv[1] )); - } - arg2 = reinterpret_cast< double * >(argp2); - svm_get_model_coefs(arg1,arg2); - return Qnil; -fail: - return Qnil; -} - - -SWIGINTERN VALUE -_wrap_svm_get_model_perm(int argc, VALUE *argv, VALUE self) { - svm_model *arg1 = (svm_model *) 0 ; - int *arg2 = (int *) 0 ; - void *argp1 = 0 ; - int res1 = 0 ; - void *argp2 = 0 ; - int res2 = 0 ; - - if ((argc < 2) || (argc > 2)) { - rb_raise(rb_eArgError, "wrong # of arguments(%d for 2)",argc); SWIG_fail; - } - res1 = SWIG_ConvertPtr(argv[0], &argp1,SWIGTYPE_p_svm_model, 0 | 0 ); - if (!SWIG_IsOK(res1)) { - SWIG_exception_fail(SWIG_ArgError(res1), Ruby_Format_TypeError( "", "svm_model *","svm_get_model_perm", 1, argv[0] )); - } - arg1 = reinterpret_cast< svm_model * >(argp1); - res2 = SWIG_ConvertPtr(argv[1], &argp2,SWIGTYPE_p_int, 0 | 0 ); - if (!SWIG_IsOK(res2)) { - SWIG_exception_fail(SWIG_ArgError(res2), Ruby_Format_TypeError( "", "int *","svm_get_model_perm", 2, argv[1] )); - } - arg2 = reinterpret_cast< int * >(argp2); - svm_get_model_perm(arg1,arg2); - return Qnil; -fail: - return Qnil; -} - - -SWIGINTERN VALUE _wrap_new_int(int argc, VALUE *argv, VALUE self) { size_t arg1 ; size_t val1 ; @@ -4513,7 +4375,6 @@ SWIGEXPORT void Init_libsvm(void) { rb_define_module_function(mLibsvm, "svm_cross_validation", VALUEFUNC(_wrap_svm_cross_validation), -1); rb_define_module_function(mLibsvm, "svm_save_model", VALUEFUNC(_wrap_svm_save_model), -1); rb_define_module_function(mLibsvm, "svm_load_model", VALUEFUNC(_wrap_svm_load_model), -1); - rb_define_module_function(mLibsvm, "svm_get_obj", VALUEFUNC(_wrap_svm_get_obj), -1); rb_define_module_function(mLibsvm, "svm_get_svm_type", VALUEFUNC(_wrap_svm_get_svm_type), -1); rb_define_module_function(mLibsvm, "svm_get_nr_class", VALUEFUNC(_wrap_svm_get_nr_class), -1); rb_define_module_function(mLibsvm, "svm_get_labels", VALUEFUNC(_wrap_svm_get_labels), -1); @@ -4524,10 +4385,6 @@ SWIGEXPORT void Init_libsvm(void) { rb_define_module_function(mLibsvm, "svm_destroy_model", VALUEFUNC(_wrap_svm_destroy_model), -1); rb_define_module_function(mLibsvm, "svm_check_parameter", VALUEFUNC(_wrap_svm_check_parameter), -1); rb_define_module_function(mLibsvm, "svm_check_probability_model", VALUEFUNC(_wrap_svm_check_probability_model), -1); - rb_define_module_function(mLibsvm, "svm_get_model_rho", VALUEFUNC(_wrap_svm_get_model_rho), -1); - rb_define_module_function(mLibsvm, "svm_get_model_num_coefs", VALUEFUNC(_wrap_svm_get_model_num_coefs), -1); - rb_define_module_function(mLibsvm, "svm_get_model_coefs", VALUEFUNC(_wrap_svm_get_model_coefs), -1); - rb_define_module_function(mLibsvm, "svm_get_model_perm", VALUEFUNC(_wrap_svm_get_model_perm), -1); rb_define_module_function(mLibsvm, "new_int", VALUEFUNC(_wrap_new_int), -1); rb_define_module_function(mLibsvm, "delete_int", VALUEFUNC(_wrap_delete_int), -1); rb_define_module_function(mLibsvm, "int_getitem", VALUEFUNC(_wrap_int_getitem), -1); ext/libsvm_wrap.cxx @@ -6,6 +6,7 @@ #include <string.h> #include <stdarg.h> #include "svm.h" +int libsvm_version = LIBSVM_VERSION; typedef float Qfloat; typedef signed char schar; #ifndef min @@ -22,36 +23,40 @@ template <class S, class T> inline void clone(T*& dst, S* src, int n) } inline double powi(double base, int times) { - double tmp = base, ret = 1.0; + double tmp = base, ret = 1.0; - for(int t=times; t>0; t/=2) + for(int t=times; t>0; t/=2) { - if(t%2==1) ret*=tmp; - tmp = tmp * tmp; - } - return ret; + if(t%2==1) ret*=tmp; + tmp = tmp * tmp; + } + return ret; } #define INF HUGE_VAL #define TAU 1e-12 #define Malloc(type,n) (type *)malloc((n)*sizeof(type)) + +static void print_string_stdout(const char *s) +{ + fputs(s,stdout); + fflush(stdout); +} +void (*svm_print_string) (const char *) = &print_string_stdout; #if 1 int info_on = 0; static void info(const char *fmt,...) { + char buf[BUFSIZ]; va_list ap; if (info_on==1) { - va_start(ap,fmt); - vprintf(fmt,ap); - va_end(ap); + va_start(ap,fmt); + vsprintf(buf,fmt,ap); + va_end(ap); + (*svm_print_string)(buf); } } -static void info_flush() -{ - if (info_on==1) fflush(stdout); -} #else -static void info(char *fmt,...) {} -static void info_flush() {} +static void info(const char *fmt,...) {} #endif // @@ -564,7 +569,7 @@ void Solver::Solve(int l, const QMatrix& Q, const double *p_, const schar *y_, { counter = min(l,1000); if(shrinking) do_shrinking(); - info("."); info_flush(); + info("."); } int i,j; @@ -574,7 +579,7 @@ void Solver::Solve(int l, const QMatrix& Q, const double *p_, const schar *y_, reconstruct_gradient(); // reset active set size and check active_size = l; - info("*"); info_flush(); + info("*"); if(select_working_set(i,j)!=0) break; else @@ -927,7 +932,7 @@ void Solver::do_shrinking() unshrink = true; reconstruct_gradient(); active_size = l; - info("*"); info_flush(); + info("*"); } for(i=0;i<active_size;i++) @@ -1110,7 +1115,7 @@ int Solver_NU::select_working_set(int &out_i, int &out_j) } if(max(Gmaxp+Gmaxp2,Gmaxn+Gmaxn2) < eps) - return 1; + return 1; if (y[Gmin_idx] == +1) out_i = Gmaxp_idx; @@ -1388,7 +1393,7 @@ public: Qfloat *buf = buffer[next_buffer]; next_buffer = 1 - next_buffer; schar si = sign[i]; - for(int j=0;j<len;j++) + for(j=0;j<len;j++) buf[j] = (Qfloat) si * (Qfloat) sign[j] * data[index[j]]; return buf; } @@ -1625,7 +1630,6 @@ struct decision_function { double *alpha; double rho; - double obj; }; decision_function svm_train_one( @@ -1682,7 +1686,6 @@ decision_function svm_train_one( decision_function f; f.alpha = alpha; f.rho = si.rho; - f.obj = si.obj; return f; } @@ -1697,11 +1700,9 @@ struct svm_model svm_node **SV; // SVs (SV[l]) double **sv_coef; // coefficients for SVs in decision functions (sv_coef[k-1][l]) double *rho; // constants in decision functions (rho[k*(k-1)/2]) - double *probA; // pariwise probability information + double *probA; // pariwise probability information double *probB; - double *obj; - // for classification only int *label; // label of each class (label[k]) @@ -1712,48 +1713,6 @@ struct svm_model // 0 if svm_model is created by svm_train }; -// Get the rho element of the model. Only works when there are two -// labels (1 classifier). -double svm_get_model_rho(struct svm_model *model) -{ - if (model->nr_class > 2) - info("warning: rho requested for model with more than 2 labels"); - return model->rho[0]; -} - -int svm_get_model_num_coefs(struct svm_model *model) -{ - return model->l; -} - -// Get the coefficients of the model. Only works when there are two -// labels (1 classifier). -void svm_get_model_coefs(struct svm_model *model, double* out_array) -{ - if (model->nr_class > 2) - info("warning: coefficients requested for model with more than 2 labels"); - memcpy(out_array, model->sv_coef[0], sizeof(double) * model->l); -} - -// Get the permutation of the indices of the coefficients w.r.t. to the input problem. -void svm_get_model_perm(struct svm_model *model, int* out_array) -{ - if (model->nr_class > 2) - info("warning: permutation requested for model with more than 2 labels"); - int i; - for ( i = 0; i < model->l; ++i) - { - struct svm_node* n = model->SV[i]; - if (0 != n->index) - { - info("warning: missing 0 index"); - out_array[i] = -1; - } - else - out_array[i] = (int)n->value; - } -} - // Platt's binary SVM Probablistic Output: an improvement from Lin et al. void sigmoid_train( int l, const double *dec_values, const double *labels, @@ -1766,7 +1725,7 @@ void sigmoid_train( if (labels[i] > 0) prior1+=1; else prior0+=1; - int max_iter=100; // Maximal number of iterations + int max_iter=100; // Maximal number of iterations double min_step=1e-10; // Minimal step taken in line search double sigma=1e-12; // For numerically strict PD of Hessian double eps=1e-5; @@ -1830,7 +1789,7 @@ void sigmoid_train( gd=g1*dA+g2*dB; - stepsize = 1; // Line Search + stepsize = 1; // Line Search while (stepsize >= min_step) { newA = A + stepsize * dA; @@ -2050,10 +2009,10 @@ double svm_svr_probability( int count=0; mae=0; for(i=0;i<prob->l;i++) - if (fabs(ymv[i]) > 5*std) - count=count+1; + if (fabs(ymv[i]) > 5*std) + count=count+1; else - mae+=fabs(ymv[i]); + mae+=fabs(ymv[i]); mae /= (prob->l-count); info("Prob. model for test data: target value = predicted value + z,\nz: Laplace distribution e^(-|z|/sigma)/(2sigma),sigma= %g\n",mae); free(ymv); @@ -2151,8 +2110,6 @@ svm_model *svm_train(const svm_problem *prob, const svm_parameter *param) decision_function f = svm_train_one(prob,param,0,0); model->rho = Malloc(double,1); model->rho[0] = f.rho; - model->obj = Malloc(double,1); - model->obj[0] = f.obj; int nSV = 0; int i; @@ -2266,13 +2223,8 @@ svm_model *svm_train(const svm_problem *prob, const svm_parameter *param) model->label[i] = label[i]; model->rho = Malloc(double,nr_class*(nr_class-1)/2); - model->obj = Malloc(double,nr_class*(nr_class-1)/2); for(i=0;i<nr_class*(nr_class-1)/2;i++) - { model->rho[i] = f[i].rho; - model->obj[i] = f[i].obj; - } - if(param->probability) { @@ -2482,10 +2434,6 @@ void svm_cross_validation(const svm_problem *prob, const svm_parameter *param, i free(perm); } -double svm_get_obj(const svm_model *model, const int i) -{ - return model->obj[i]; -} int svm_get_svm_type(const svm_model *model) { @@ -2511,7 +2459,7 @@ double svm_get_svr_probability(const svm_model *model) return model->probA[0]; else { - info("Model doesn't contain information for SVR probability inference\n"); + fprintf(stderr,"Model doesn't contain information for SVR probability inference\n"); return 0; } } @@ -2647,7 +2595,7 @@ double svm_predict_probability( for(i=0;i<nr_class;i++) free(pairwise_prob[i]); free(dec_values); - free(pairwise_prob); + free(pairwise_prob); return model->label[prob_max_idx]; } else @@ -2751,6 +2699,27 @@ int svm_save_model(const char *model_file_name, const svm_model *model) else return 0; } +static char *line = NULL; +static int max_line_len; + +static char* readline(FILE *input) +{ + int len; + + if(fgets(line,max_line_len,input) == NULL) + return NULL; + + while(strrchr(line,'\n') == NULL) + { + max_line_len *= 2; + line = (char *) realloc(line,max_line_len); + len = (int) strlen(line); + if(fgets(line+len,max_line_len-len,input) == NULL) + break; + } + return line; +} + svm_model *svm_load_model(const char *model_file_name) { FILE *fp = fopen(model_file_name,"rb"); @@ -2766,8 +2735,6 @@ svm_model *svm_load_model(const char *model_file_name) model->label = NULL; model->nSV = NULL; - model->obj = NULL; - char cmd[81]; while(1) { @@ -2789,7 +2756,6 @@ svm_model *svm_load_model(const char *model_file_name) { fprintf(stderr,"unknown svm type.\n"); free(model->rho); - free(model->obj); free(model->label); free(model->nSV); free(model); @@ -2812,7 +2778,6 @@ svm_model *svm_load_model(const char *model_file_name) { fprintf(stderr,"unknown kernel function.\n"); free(model->rho); - free(model->obj); free(model->label); free(model->nSV); free(model); @@ -2877,7 +2842,6 @@ svm_model *svm_load_model(const char *model_file_name) { fprintf(stderr,"unknown text in model file: [%s]\n",cmd); free(model->rho); - free(model->obj); free(model->label); free(model->nSV); free(model); @@ -2890,23 +2854,23 @@ svm_model *svm_load_model(const char *model_file_name) int elements = 0; long pos = ftell(fp); - while(1) + max_line_len = 1024; + line = Malloc(char,max_line_len); + char *p,*endptr,*idx,*val; + + while(readline(fp)!=NULL) { - int c = fgetc(fp); - switch(c) + p = strtok(line,":"); + while(1) { - case '\n': - // count the '-1' element - case ':': - ++elements; + p = strtok(NULL,":"); + if(p == NULL) break; - case EOF: - goto out; - default: - ; + ++elements; } } -out: + elements += model->l; + fseek(fp,pos,SEEK_SET); int m = model->nr_class - 1; @@ -2916,30 +2880,41 @@ out: for(i=0;i<m;i++) model->sv_coef[i] = Malloc(double,l); model->SV = Malloc(svm_node*,l); - svm_node *x_space=NULL; + svm_node *x_space = NULL; if(l>0) x_space = Malloc(svm_node,elements); int j=0; for(i=0;i<l;i++) { + readline(fp); model->SV[i] = &x_space[j]; - for(int k=0;k<m;k++) - fscanf(fp,"%lf",&model->sv_coef[k][i]); + + p = strtok(line, " \t"); + model->sv_coef[0][i] = strtod(p,&endptr); + for(int k=1;k<m;k++) + { + p = strtok(NULL, " \t"); + model->sv_coef[k][i] = strtod(p,&endptr); + } + while(1) { - int c; - do { - c = getc(fp); - if(c=='\n') goto out2; - } while(isspace(c)); - ungetc(c,fp); - fscanf(fp,"%d:%lf",&(x_space[j].index),&(x_space[j].value)); + idx = strtok(NULL, ":"); + val = strtok(NULL, " \t"); + + if(val == NULL) + break; + x_space[j].index = (int) strtol(idx,&endptr,10); + x_space[j].value = strtod(val,&endptr); + ++j; - } -out2: + } x_space[j++].index = -1; } - if (ferror(fp) != 0 || fclose(fp) != 0) return NULL; + free(line); + + if (ferror(fp) != 0 || fclose(fp) != 0) + return NULL; model->free_sv = 1; // XXX return model; @@ -2954,7 +2929,6 @@ void svm_destroy_model(svm_model* model) free(model->SV); free(model->sv_coef); free(model->rho); - free(model->obj); free(model->label); free(model->probA); free(model->probB); ext/svm.cpp @@ -1,12 +1,14 @@ #ifndef _LIBSVM_H #define _LIBSVM_H -#define LIBSVM_VERSION 288 +#define LIBSVM_VERSION 289 #ifdef __cplusplus extern "C" { #endif +extern int libsvm_version; + struct svm_node { int index; @@ -50,8 +52,6 @@ void svm_cross_validation(const struct svm_problem *prob, const struct svm_param int svm_save_model(const char *model_file_name, const struct svm_model *model); struct svm_model *svm_load_model(const char *model_file_name); -double svm_get_obj(const struct svm_model *model, const int i); - int svm_get_svm_type(const struct svm_model *model); int svm_get_nr_class(const struct svm_model *model); void svm_get_labels(const struct svm_model *model, int *label); @@ -67,12 +67,7 @@ void svm_destroy_param(struct svm_parameter *param); const char *svm_check_parameter(const struct svm_problem *prob, const struct svm_parameter *param); int svm_check_probability_model(const struct svm_model *model); -double svm_get_model_rho(struct svm_model *model); -int svm_get_model_num_coefs(struct svm_model *model); -// out_array must have svm_get_model_num_coefs(model) elements -void svm_get_model_coefs(struct svm_model *model, double* out_array); -// out_array must have svm_get_model_num_coef(model)s elements -void svm_get_model_perm(struct svm_model *model, int* out_array); +extern void (*svm_print_string) (const char *); #ifdef __cplusplus } ext/svm.h 5b3ec457c1aaa866b8d7890a37e8c4f36e7f4803 dev tom.z.zeng@gmail.com http://github.com/tomz/libsvm-ruby-swig/commit/a88656c16fb178bc867e4611b0ae2bc7c36394df a88656c16fb178bc867e4611b0ae2bc7c36394df 2009-05-15T13:46:20-07:00 2009-05-15T13:46:20-07:00 Sync with LIBSVM 2.89 files c36e7857fb9cf9fc5e999720f63c22ae212366f2 dev tom.z.zeng@gmail.com