Add Normalisation Option

linear.cpp

@@ -2620,6 +2620,8 @@ int save_model(const char *model_file_name, const struct model *model_)
 
 	fprintf(fp, "bias %.16g\n", model_->bias);
 
+	fprintf(fp, "normalization %d\n", model_->normal);
+
 	fprintf(fp, "w\n");
 	for(i=0; i<w_size; i++)
 	{
@@ -2645,6 +2647,7 @@ struct model *load_model(const char *model_file_name)
 	int nr_feature;
 	int n;
 	int nr_class;
+	int normalfac;
 	double bias;
 	model *model_ = Malloc(model,1);
 	parameter& param = model_->param;
@@ -2686,6 +2689,11 @@ struct model *load_model(const char *model_file_name)
 			fscanf(fp,"%d",&nr_class);
 			model_->nr_class=nr_class;
 		}
+		else if(strcmp(cmd,"normalization")==0)
+		{
+			fscanf(fp,"%d",&normalfac);
+			model_->normal=normalfac;
+		}
 		else if(strcmp(cmd,"nr_feature")==0)
 		{
 			fscanf(fp,"%d",&nr_feature);

linear.h

@@ -39,6 +39,7 @@ struct model
 	struct parameter param;
 	int nr_class;		/* number of classes */
 	int nr_feature;
+	int normal;
 	double *w;
 	int *label;		/* label of each class */
 	double bias;

predict.c

@@ -1,6 +1,7 @@
 #include <stdio.h>
 #include <ctype.h>
 #include <stdlib.h>
+#include <math.h>
 #include <string.h>
 #include <errno.h>
 #include "linear.h"
@@ -133,6 +134,17 @@ void do_predict(FILE *input, FILE *output)
 		}
 		x[i].index = -1;
 
+		if(model_->normal){
+			double length = 0;
+			for(int kk = 0; x[kk].index != -1; kk++)
+				length += x[kk].value * x[kk].value;
+
+			length = sqrt(length);
+
+			for(int kk = 0; x[kk].index != -1; kk++)
+				x[kk].value /= length;
+		}
+
 		if(flag_predict_probability)
 		{
 			int j;

train.c

@@ -93,6 +93,24 @@ int flag_cross_validation;
 int nr_fold;
 double bias;
 
+int normal = 0;
+
+void normalize(problem *pb){
+	feature_node **p = pb->x;
+
+	for(int i = 0; i < pb->l; i++){
+		double length = 0;
+
+		for(int j = 0; p[i][j].index != -1; j++)
+			length += (p[i][j].value) * (p[i][j].value);
+
+		length = sqrt(length);
+
+		for(int j = 0; p[i][j].index != -1; j++)
+			p[i][j].value /= length;
+	}
+}
+
 int main(int argc, char **argv)
 {
 	char input_file_name[1024];
@@ -101,6 +119,7 @@ int main(int argc, char **argv)
 
 	parse_command_line(argc, argv, input_file_name, model_file_name);
 	read_problem(input_file_name);
+	if(normal) normalize(&prob);
 	error_msg = check_parameter(&prob,&param);
 
 	if(error_msg)
@@ -116,6 +135,7 @@ int main(int argc, char **argv)
 	else
 	{
 		model_=train(&prob, &param);
+		model_->normal = normal;
 		if(save_model(model_file_name, model_))
 		{
 			fprintf(stderr,"can't save model to file %s\n",model_file_name);
@@ -197,6 +217,11 @@ void parse_command_line(int argc, char **argv, char *input_file_name, char *mode
 			exit_with_help();
 		switch(argv[i-1][1])
 		{
+			case 'n':
+				normal = 1;
+				i--;
+				break;
+
 			case 's':
 				param.solver_type = atoi(argv[i]);
 				break;
@@ -252,7 +277,7 @@ void parse_command_line(int argc, char **argv, char *input_file_name, char *mode
 	// determine filenames
 	if(i>=argc)
 		exit_with_help();
-
+	
 	strcpy(input_file_name, argv[i]);
 
 	if(i<argc-1)