liblinear-ruby 1.0.1 → 1.0.2

data/ext/linear.cpp

@@ -6,6 +6,7 @@
 #include <locale.h>
 #include "linear.h"
 #include "tron.h"
+int liblinear_version = LIBLINEAR_VERSION;
 typedef signed char schar;
 template <class T> static inline void swap(T& x, T& y) { T t=x; x=y; y=t; }
 #ifndef min
@@ -173,7 +174,6 @@ void l2r_lr_fun::Hv(double *s, double *Hs)
 	int i;
 	int l=prob->l;
 	int w_size=get_nr_variable();
-	double *wa = new double[l];
 	feature_node **x=prob->x;
 
 	for(i=0;i<w_size;i++)
@@ -181,15 +181,14 @@ void l2r_lr_fun::Hv(double *s, double *Hs)
 	for(i=0;i<l;i++)
 	{
 		feature_node * const xi=x[i];
-		wa[i] = sparse_operator::dot(s, xi);
-		
-		wa[i] = C[i]*D[i]*wa[i];
+		double xTs = sparse_operator::dot(s, xi);
 
-		sparse_operator::axpy(wa[i], xi, Hs);
+		xTs = C[i]*D[i]*xTs;
+
+		sparse_operator::axpy(xTs, xi, Hs);
 	}
 	for(i=0;i<w_size;i++)
 		Hs[i] = s[i] + Hs[i];
-	delete[] wa;
 }
 
 void l2r_lr_fun::Xv(double *v, double *Xv)
@@ -233,7 +232,6 @@ protected:
 
 	double *C;
 	double *z;
-	double *D;
 	int *I;
 	int sizeI;
 	const problem *prob;
@@ -246,7 +244,6 @@ l2r_l2_svc_fun::l2r_l2_svc_fun(const problem *prob, double *C)
 	this->prob = prob;
 
 	z = new double[l];
-	D = new double[l];
 	I = new int[l];
 	this->C = C;
 }
@@ -254,7 +251,6 @@ l2r_l2_svc_fun::l2r_l2_svc_fun(const problem *prob, double *C)
 l2r_l2_svc_fun::~l2r_l2_svc_fun()
 {
 	delete[] z;
-	delete[] D;
 	delete[] I;
 }
 
@@ -312,7 +308,6 @@ void l2r_l2_svc_fun::Hv(double *s, double *Hs)
 {
 	int i;
 	int w_size=get_nr_variable();
-	double *wa = new double[sizeI];
 	feature_node **x=prob->x;
 
 	for(i=0;i<w_size;i++)
@@ -320,15 +315,14 @@ void l2r_l2_svc_fun::Hv(double *s, double *Hs)
 	for(i=0;i<sizeI;i++)
 	{
 		feature_node * const xi=x[I[i]];
-		wa[i] = sparse_operator::dot(s, xi);
-		
-		wa[i] = C[I[i]]*wa[i];
+		double xTs = sparse_operator::dot(s, xi);
+
+		xTs = C[I[i]]*xTs;
 
-		sparse_operator::axpy(wa[i], xi, Hs);
+		sparse_operator::axpy(xTs, xi, Hs);
 	}
 	for(i=0;i<w_size;i++)
 		Hs[i] = s[i] + 2*Hs[i];
-	delete[] wa;
 }
 
 void l2r_l2_svc_fun::Xv(double *v, double *Xv)
@@ -431,19 +425,19 @@ void l2r_l2_svr_fun::grad(double *w, double *g)
 		g[i] = w[i] + 2*g[i];
 }
 
-// A coordinate descent algorithm for 
+// A coordinate descent algorithm for
 // multi-class support vector machines by Crammer and Singer
 //
 //  min_{\alpha}  0.5 \sum_m ||w_m(\alpha)||^2 + \sum_i \sum_m e^m_i alpha^m_i
 //    s.t.     \alpha^m_i <= C^m_i \forall m,i , \sum_m \alpha^m_i=0 \forall i
-// 
+//
 //  where e^m_i = 0 if y_i  = m,
 //        e^m_i = 1 if y_i != m,
-//  C^m_i = C if m  = y_i, 
-//  C^m_i = 0 if m != y_i, 
-//  and w_m(\alpha) = \sum_i \alpha^m_i x_i 
+//  C^m_i = C if m  = y_i,
+//  C^m_i = 0 if m != y_i,
+//  and w_m(\alpha) = \sum_i \alpha^m_i x_i
 //
-// Given: 
+// Given:
 // x, y, C
 // eps is the stopping tolerance
 //
@@ -551,7 +545,7 @@ void Solver_MCSVM_CS::Solve(double *w)
 	double eps_shrink = max(10.0*eps, 1.0); // stopping tolerance for shrinking
 	bool start_from_all = true;
 
-	// Initial alpha can be set here. Note that 
+	// Initial alpha can be set here. Note that
 	// sum_m alpha[i*nr_class+m] = 0, for all i=1,...,l-1
 	// alpha[i*nr_class+m] <= C[GETI(i)] if prob->y[i] == m
 	// alpha[i*nr_class+m] <= 0 if prob->y[i] != m
@@ -746,14 +740,14 @@ void Solver_MCSVM_CS::Solve(double *w)
 	delete [] active_size_i;
 }
 
-// A coordinate descent algorithm for 
+// A coordinate descent algorithm for
 // L1-loss and L2-loss SVM dual problems
 //
 //  min_\alpha  0.5(\alpha^T (Q + D)\alpha) - e^T \alpha,
 //    s.t.      0 <= \alpha_i <= upper_bound_i,
-// 
+//
 //  where Qij = yi yj xi^T xj and
-//  D is a diagonal matrix 
+//  D is a diagonal matrix
 //
 // In L1-SVM case:
 // 		upper_bound_i = Cp if y_i = 1
@@ -764,12 +758,12 @@ void Solver_MCSVM_CS::Solve(double *w)
 // 		D_ii = 1/(2*Cp)	if y_i = 1
 // 		D_ii = 1/(2*Cn)	if y_i = -1
 //
-// Given: 
+// Given:
 // x, y, Cp, Cn
 // eps is the stopping tolerance
 //
 // solution will be put in w
-// 
+//
 // See Algorithm 3 of Hsieh et al., ICML 2008
 
 #undef GETI
@@ -951,14 +945,14 @@ static void solve_l2r_l1l2_svc(
 }
 
 
-// A coordinate descent algorithm for 
+// A coordinate descent algorithm for
 // L1-loss and L2-loss epsilon-SVR dual problem
 //
 //  min_\beta  0.5\beta^T (Q + diag(lambda)) \beta - p \sum_{i=1}^l|\beta_i| + \sum_{i=1}^l yi\beta_i,
 //    s.t.      -upper_bound_i <= \beta_i <= upper_bound_i,
-// 
+//
 //  where Qij = xi^T xj and
-//  D is a diagonal matrix 
+//  D is a diagonal matrix
 //
 // In L1-SVM case:
 // 		upper_bound_i = C
@@ -967,13 +961,13 @@ static void solve_l2r_l1l2_svc(
 // 		upper_bound_i = INF
 // 		lambda_i = 1/(2*C)
 //
-// Given: 
+// Given:
 // x, y, p, C
 // eps is the stopping tolerance
 //
 // solution will be put in w
 //
-// See Algorithm 4 of Ho and Lin, 2012   
+// See Algorithm 4 of Ho and Lin, 2012
 
 #undef GETI
 #define GETI(i) (0)
@@ -1165,17 +1159,17 @@ static void solve_l2r_l1l2_svr(
 }
 
 
-// A coordinate descent algorithm for 
+// A coordinate descent algorithm for
 // the dual of L2-regularized logistic regression problems
 //
 //  min_\alpha  0.5(\alpha^T Q \alpha) + \sum \alpha_i log (\alpha_i) + (upper_bound_i - \alpha_i) log (upper_bound_i - \alpha_i),
 //    s.t.      0 <= \alpha_i <= upper_bound_i,
-// 
-//  where Qij = yi yj xi^T xj and 
+//
+//  where Qij = yi yj xi^T xj and
 //  upper_bound_i = Cp if y_i = 1
 //  upper_bound_i = Cn if y_i = -1
 //
-// Given: 
+// Given:
 // x, y, Cp, Cn
 // eps is the stopping tolerance
 //
@@ -1194,7 +1188,7 @@ void solve_l2r_lr_dual(const problem *prob, double *w, double eps, double Cp, do
 	int i, s, iter = 0;
 	double *xTx = new double[l];
 	int max_iter = 1000;
-	int *index = new int[l];	
+	int *index = new int[l];
 	double *alpha = new double[2*l]; // store alpha and C - alpha
 	schar *y = new schar[l];
 	int max_inner_iter = 100; // for inner Newton
@@ -1213,7 +1207,7 @@ void solve_l2r_lr_dual(const problem *prob, double *w, double eps, double Cp, do
 			y[i] = -1;
 		}
 	}
-	
+
 	// Initial alpha can be set here. Note that
 	// 0 < alpha[i] < upper_bound[GETI(i)]
 	// alpha[2*i] + alpha[2*i+1] = upper_bound[GETI(i)]
@@ -1328,12 +1322,12 @@ void solve_l2r_lr_dual(const problem *prob, double *w, double eps, double Cp, do
 	delete [] index;
 }
 
-// A coordinate descent algorithm for 
+// A coordinate descent algorithm for
 // L1-regularized L2-loss support vector classification
 //
 //  min_w \sum |wj| + C \sum max(0, 1-yi w^T xi)^2,
 //
-// Given: 
+// Given:
 // x, y, Cp, Cn
 // eps is the stopping tolerance
 //
@@ -1607,12 +1601,12 @@ static void solve_l1r_l2_svc(
 	delete [] xj_sq;
 }
 
-// A coordinate descent algorithm for 
+// A coordinate descent algorithm for
 // L1-regularized logistic regression problems
 //
 //  min_w \sum |wj| + C \sum log(1+exp(-yi w^T xi)),
 //
-// Given: 
+// Given:
 // x, y, Cp, Cn
 // eps is the stopping tolerance
 //
@@ -2075,8 +2069,8 @@ static void group_classes(const problem *prob, int *nr_class_ret, int **label_re
 	}
 
 	//
-	// Labels are ordered by their first occurrence in the training set. 
-	// However, for two-class sets with -1/+1 labels and -1 appears first, 
+	// Labels are ordered by their first occurrence in the training set.
+	// However, for two-class sets with -1/+1 labels and -1 appears first,
 	// we swap labels to ensure that internally the binary SVM has positive data corresponding to the +1 instances.
 	//
 	if (nr_class == 2 && label[0] == -1 && label[1] == 1)
@@ -2350,7 +2344,7 @@ model* train(const problem *prob, const parameter *param)
 					sub_prob.y[k] = +1;
 				for(; k<sub_prob.l; k++)
 					sub_prob.y[k] = -1;
-				
+
 				if(param->init_sol != NULL)
 					for(i=0;i<w_size;i++)
 						model_->w[i] = param->init_sol[i];
@@ -2386,7 +2380,7 @@ model* train(const problem *prob, const parameter *param)
 
 					train_one(&sub_prob, param, w, weighted_C[i], param->C);
 
-					for(int j=0;j<w_size;j++)
+					for(j=0;j<w_size;j++)
 						model_->w[j*nr_class+i] = w[j];
 				}
 				free(w);
@@ -2601,7 +2595,7 @@ void find_parameter_C(const problem *prob, const parameter *param, int nr_fold,
 		param1.C = param1.C*ratio;
 	}
 
-	if(param1.C > max_C && max_C > start_C) 
+	if(param1.C > max_C && max_C > start_C)
 		info("warning: maximum C reached.\n");
 	free(fold_start);
 	free(perm);
@@ -2891,11 +2885,6 @@ struct model *load_model(const char *model_file_name)
 		int j;
 		for(j=0; j<nr_w; j++)
 			FSCANF(fp, "%lf ", &model_->w[i*nr_w+j]);
-		if (fscanf(fp, "\n") !=0)
-		{
-			fprintf(stderr, "ERROR: fscanf failed to read the model\n");
-			EXIT_LOAD_MODEL()
-		}
 	}
 
 	setlocale(LC_ALL, old_locale);
@@ -2924,7 +2913,7 @@ void get_labels(const model *model_, int* label)
 }
 
 // use inline here for better performance (around 20% faster than the non-inline one)
-static inline double get_w_value(const struct model *model_, int idx, int label_idx) 
+static inline double get_w_value(const struct model *model_, int idx, int label_idx)
 {
 	int nr_class = model_->nr_class;
 	int solver_type = model_->param.solver_type;
@@ -2934,7 +2923,7 @@ static inline double get_w_value(const struct model *model_, int idx, int label_
 		return 0;
 	if(check_regression_model(model_))
 		return w[idx];
-	else 
+	else
 	{
 		if(label_idx < 0 || label_idx >= nr_class)
 			return 0;
@@ -3022,7 +3011,7 @@ const char *check_parameter(const problem *prob, const parameter *param)
 		&& param->solver_type != L2R_L1LOSS_SVR_DUAL)
 		return "unknown solver type";
 
-	if(param->init_sol != NULL 
+	if(param->init_sol != NULL
 		&& param->solver_type != L2R_LR && param->solver_type != L2R_L2LOSS_SVC)
 		return "Initial-solution specification supported only for solver L2R_LR and L2R_L2LOSS_SVC";
 

data/ext/linear.h

@@ -1,10 +1,14 @@
 #ifndef _LIBLINEAR_H
 #define _LIBLINEAR_H
 
+#define LIBLINEAR_VERSION 211
+
 #ifdef __cplusplus
 extern "C" {
 #endif
 
+extern int liblinear_version;
+
 struct feature_node
 {
 	int index;
@@ -16,7 +20,7 @@ struct problem
 	int l, n;
 	double *y;
 	struct feature_node **x;
-	double bias;            /* < 0 if no bias term */  
+	double bias;            /* < 0 if no bias term */
 };
 
 enum { L2R_LR, L2R_L2LOSS_SVC_DUAL, L2R_L2LOSS_SVC, L2R_L1LOSS_SVC_DUAL, MCSVM_CS, L1R_L2LOSS_SVC, L1R_LR, L2R_LR_DUAL, L2R_L2LOSS_SVR = 11, L2R_L2LOSS_SVR_DUAL, L2R_L1LOSS_SVR_DUAL }; /* solver_type */

data/ext/tron.cpp

@@ -91,9 +91,10 @@ void TRON::tron(double *w)
 	iter = 1;
 
 	double *w_new = new double[n];
+	bool reach_boundary;
 	while (iter <= max_iter && search)
 	{
-		cg_iter = trcg(delta, g, s, r);
+		cg_iter = trcg(delta, g, s, r, &reach_boundary);
 
 		memcpy(w_new, w, sizeof(double)*n);
 		daxpy_(&n, &one, s, &inc, w_new, &inc);
@@ -124,7 +125,12 @@ void TRON::tron(double *w)
 		else if (actred < eta2*prered)
 			delta = max(sigma1*delta, min(alpha*snorm, sigma3*delta));
 		else
-			delta = max(delta, min(alpha*snorm, sigma3*delta));
+		{
+			if (reach_boundary)
+				delta = sigma3*delta;
+			else
+				delta = max(delta, min(alpha*snorm, sigma3*delta));
+		}
 
 		info("iter %2d act %5.3e pre %5.3e delta %5.3e f %5.3e |g| %5.3e CG %3d\n", iter, actred, prered, delta, f, gnorm, cg_iter);
 
@@ -144,9 +150,9 @@ void TRON::tron(double *w)
 			info("WARNING: f < -1.0e+32\n");
 			break;
 		}
-		if (fabs(actred) <= 0 && prered <= 0)
+		if (prered <= 0)
 		{
-			info("WARNING: actred and prered <= 0\n");
+			info("WARNING: prered <= 0\n");
 			break;
 		}
 		if (fabs(actred) <= 1.0e-12*fabs(f) &&
@@ -163,7 +169,7 @@ void TRON::tron(double *w)
 	delete[] s;
 }
 
-int TRON::trcg(double delta, double *g, double *s, double *r)
+int TRON::trcg(double delta, double *g, double *s, double *r, bool *reach_boundary)
 {
 	int i, inc = 1;
 	int n = fun_obj->get_nr_variable();
@@ -172,6 +178,7 @@ int TRON::trcg(double delta, double *g, double *s, double *r)
 	double *Hd = new double[n];
 	double rTr, rnewTrnew, alpha, beta, cgtol;
 
+	*reach_boundary = false;
 	for (i=0; i<n; i++)
 	{
 		s[i] = 0;
@@ -194,6 +201,7 @@ int TRON::trcg(double delta, double *g, double *s, double *r)
 		if (dnrm2_(&n, s, &inc) > delta)
 		{
 			info("cg reaches trust region boundary\n");
+			*reach_boundary = true;
 			alpha = -alpha;
 			daxpy_(&n, &alpha, d, &inc, s, &inc);
 

data/ext/tron.h

@@ -22,7 +22,7 @@ public:
 	void set_print_string(void (*i_print) (const char *buf));
 
 private:
-	int trcg(double delta, double *g, double *s, double *r);
+	int trcg(double delta, double *g, double *s, double *r, bool *reach_boundary);
 	double norm_inf(int n, double *x);
 
 	double eps;

data/lib/liblinear.rb

@@ -26,6 +26,8 @@ class Liblinear
   L2R_L2LOSS_SVR_DUAL = Liblinearswig::L2R_L2LOSS_SVR_DUAL
   L2R_L1LOSS_SVR_DUAL = Liblinearswig::L2R_L1LOSS_SVR_DUAL
 
+  LIBLINEAR_VERSION   = sprintf("%.2f", Liblinearswig::LIBLINEAR_VERSION / 100.0)
+
   class << self
     # @param problem [Liblinear::Problem]
     # @param parameter [Liblinear::Parameter]

data/lib/liblinear/version.rb

@@ -1,3 +1,3 @@
 class Liblinear
-  VERSION = '1.0.1'
+  VERSION = '1.0.2'
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: liblinear-ruby
 version: !ruby/object:Gem::Version
-  version: 1.0.1
+  version: 1.0.2
 platform: ruby
 authors:
 - Kei Tsuchiya
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2017-02-08 00:00:00.000000000 Z
+date: 2017-05-22 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler
@@ -76,45 +76,6 @@ files:
 - lib/liblinear/parameter.rb
 - lib/liblinear/problem.rb
 - lib/liblinear/version.rb
-- liblinear-2.1/COPYRIGHT
-- liblinear-2.1/Makefile
-- liblinear-2.1/Makefile.win
-- liblinear-2.1/README
-- liblinear-2.1/blas/Makefile
-- liblinear-2.1/blas/blas.h
-- liblinear-2.1/blas/blasp.h
-- liblinear-2.1/blas/daxpy.c
-- liblinear-2.1/blas/ddot.c
-- liblinear-2.1/blas/dnrm2.c
-- liblinear-2.1/blas/dscal.c
-- liblinear-2.1/heart_scale
-- liblinear-2.1/linear.cpp
-- liblinear-2.1/linear.def
-- liblinear-2.1/linear.h
-- liblinear-2.1/matlab/Makefile
-- liblinear-2.1/matlab/README
-- liblinear-2.1/matlab/libsvmread.c
-- liblinear-2.1/matlab/libsvmwrite.c
-- liblinear-2.1/matlab/linear_model_matlab.c
-- liblinear-2.1/matlab/linear_model_matlab.h
-- liblinear-2.1/matlab/make.m
-- liblinear-2.1/matlab/predict.c
-- liblinear-2.1/matlab/train.c
-- liblinear-2.1/predict.c
-- liblinear-2.1/python/Makefile
-- liblinear-2.1/python/README
-- liblinear-2.1/python/liblinear.py
-- liblinear-2.1/python/liblinearutil.py
-- liblinear-2.1/train.c
-- liblinear-2.1/tron.cpp
-- liblinear-2.1/tron.h
-- liblinear-2.1/windows/liblinear.dll
-- liblinear-2.1/windows/libsvmread.mexw64
-- liblinear-2.1/windows/libsvmwrite.mexw64
-- liblinear-2.1/windows/predict.exe
-- liblinear-2.1/windows/predict.mexw64
-- liblinear-2.1/windows/train.exe
-- liblinear-2.1/windows/train.mexw64
 - liblinear-ruby.gemspec
 - spec/liblinear/array/double_spec.rb
 - spec/liblinear/example_spec.rb