svm.h

#ifndef _LIBSVM_H
#define _LIBSVM_H

#define LIBSVM_VERSION 290

#ifdef _WIN32

#ifndef _WINDLL
# define LIBSVM_EXPORT __declspec(dllimport)
#else
# define LIBSVM_EXPORT __declspec(dllexport)
#endif

#else
# define LIBSVM_EXPORT 
#endif /* WIN32 */

#ifdef __cplusplus
extern "C" {
#endif

extern int libsvm_version;

struct svm_node
{
	int index;
	double value;
};

/**
 * 	The problem: the dimension (l), the labels (*y) and the patterns (**x)
 * */
struct svm_problem
{
	int l;
	double *y;
	struct svm_node **x;
};

/**
 * SVM Type: Classification, Classification with L2SVM, nu-Classification, 
 * One Class Classification, epsilon Regression, nu Regression, 
 * Support Vector Domain Description (SVDD), SVDD with L2SVM
 * 
 * C_SVC_L2, SVDD and SVDD_L2 were backported from the code available at
 * 
 *    http://www.csie.ntu.edu.tw/~cjlin/libsvmtools/#18
 * 
 * written by Leland Wang and extended by Holger Froehlich.
 * 
 * --
 * Vincenzo Russo (vincenzo.russo@neminis.org)
 * */
enum { C_SVC, C_SVC_L2, NU_SVC, ONE_CLASS, EPSILON_SVR, NU_SVR, SVDD, SVDD_L2 };

/**
 * Kernel Type: Linear, Polynomial, Gaussian, Sigmoid, Stump, Perceptron, Laplace, Expo, Precomputed
 * 
 * Stump, Perceptron, Laplace, Expo were backported from the LIBSVM extension by Hsuan-Tien Lin
 * 
 * For his version of the software, see http://www.work.caltech.edu/~htlin/program/libsvm/#infensemble
 * For explanation about the four above kernels, see his Master's Thesis
 * 	"H.-T. Lin. Infinite Ensemble Learning with Support Vector Machines. 
 *   Master's Thesis, California Institute of Technology, May 2005."
 * http://www.work.caltech.edu/~htlin/publication/
 * 
 * --
 * Vincenzo Russo (vincenzo.russo@neminis.org)
 * */
enum { LINEAR, POLY, GAUSSIAN, SIGMOID, STUMP, PERC, LAPLACE, EXPO, PRECOMPUTED };	/* kernel_type */


/**
 * The parameters of a task
 * */
struct svm_parameter
{
	int svm_type;		/* SVM Type */
	int kernel_type;	/* Kernel Type */
	int degree;		/* for poly */
	double gamma;		/* for poly/gaussian/sigmoid */
	double coef0;		/* for poly/sigmoid */

	/* these are for training only */
	double cache_size; 		/* in MB */
	double eps;			/* stopping criteria (the lower eps, the better the approximation) */
	double C;			/* for C_SVC, SVDD, EPSILON_SVR and NU_SVR */
	int nr_weight;			/* for C_SVC */
	int *weight_label;		/* for C_SVC */
	double* weight;			/* for C_SVC */
	double nu;			/* for NU_SVC, ONE_CLASS, and NU_SVR */
	double p;			/* for EPSILON_SVR */
	int shrinking;			/* use the shrinking heuristics */
	int probability; 		/* do probability estimates */
};

/**
 * The model yielded by the training of an SVM
 * */
struct svm_model
{
	struct svm_parameter param;	// parameter
	int nr_class;		// number of classes, = 2 in regression/one class svm
	int l;			// total #SV
	struct svm_node **SV;	// SVs (SV[l])
	
	int *BSV_idx;		// indices of the BSVs in the original dataset
	int *SV_idx;		// indices of the SVs in the original dataset
	
	int lbsv;		// total #BSV
	
	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 *probB;

	// for classification only

	int *label;		// label of each class (label[k])
	int *nSV;		// number of SVs for each class (nSV[k])
				// nSV[0] + nSV[1] + ... + nSV[k-1] = l
	// XXX
	int free_sv;		// 1 if svm_model is created by svm_load_model
				// 0 if svm_model is created by svm_train
};


/**
 * The decision function
 * */
struct decision_function
{
	double *alpha;	// Lagrangian multipliers
	double rho;		
	
	int *BSV_idx;	// indices of the BSVs in the original dataset
	int *SV_idx;	// indices of the SVs in the original dataset
		
	int l,		// number of total SVs (SVs + BSVs) 
	    lbsv;	// number of BSVs
};


LIBSVM_EXPORT struct svm_model *svm_train(const struct svm_problem *prob, const struct svm_parameter *param);
LIBSVM_EXPORT void svm_cross_validation(const struct svm_problem *prob, const struct svm_parameter *param, int nr_fold, double *target);

LIBSVM_EXPORT int svm_save_model(const char *model_file_name, const struct svm_model *model);
LIBSVM_EXPORT struct svm_model *svm_load_model(const char *model_file_name);

LIBSVM_EXPORT int svm_get_svm_type(const struct svm_model *model);
LIBSVM_EXPORT int svm_get_nr_class(const struct svm_model *model);
LIBSVM_EXPORT void svm_get_labels(const struct svm_model *model, int *label);
LIBSVM_EXPORT double svm_get_svr_probability(const struct svm_model *model);

LIBSVM_EXPORT void svm_predict_values(const struct svm_model *model, const struct svm_node *x, double* dec_values);
LIBSVM_EXPORT double svm_predict(const struct svm_model *model, const struct svm_node *x);
LIBSVM_EXPORT double svm_predict_probability(const struct svm_model *model, const struct svm_node *x, double* prob_estimates);

LIBSVM_EXPORT void svm_destroy_model(struct svm_model *model);
LIBSVM_EXPORT void svm_destroy_param(struct svm_parameter *param);

LIBSVM_EXPORT const char *svm_check_parameter(const struct svm_problem *prob, const struct svm_parameter *param);
LIBSVM_EXPORT int svm_check_probability_model(const struct svm_model *model);

LIBSVM_EXPORT extern void (*svm_print_string) (const char *);

#ifdef __cplusplus
}
#endif

#endif /* _LIBSVM_H */