Darwin  1.10(beta)
 All Classes Namespaces Files Functions Variables Typedefs Enumerations Enumerator Friends Groups Pages
Public Member Functions | Static Public Attributes | Protected Member Functions | Protected Attributes | List of all members
drwnLinearRegressorBase Class Referenceabstract

Common functionality for drwnLinearRegressor. More...

Inheritance diagram for drwnLinearRegressorBase:
drwnRegression drwnOptimizer drwnStdObjIface drwnProperties drwnWriteable drwnCloneable drwnTypeable drwnTLinearRegressor< FeatureMap >

Public Member Functions

 drwnLinearRegressorBase ()
 default constructor
 drwnLinearRegressorBase (unsigned n)
 construct a linear regressor for data of dimension n
 drwnLinearRegressorBase (const drwnLinearRegressorBase &r)
 copy constructor
virtual const char * type () const
 returns object type as a string (e.g., Foo::type() { return "Foo"; })
virtual bool save (drwnXMLNode &xml) const
 write object to XML node (see also write)
virtual bool load (drwnXMLNode &xml)
 read object from XML node (see also read)
virtual double train (const drwnRegressionDataset &dataset)
 estimate the regression parameters a drwnRegressionDataset
virtual double getRegression (const vector< double > &features) const =0
 return the estimated value for a given feature vector
- Public Member Functions inherited from drwnRegression
 drwnRegression ()
 default constructor
 drwnRegression (unsigned n)
 construct a regression object for data of dimension n
 drwnRegression (const drwnRegression &r)
 copy constructor
int numFeatures () const
 return the dimensionality of the feature space
virtual bool valid () const
 return true if the regressor has valid parameters (i.e., has been trained)
virtual void initialize (unsigned n)
 initialize the regressor to accept data of dimensionality n
virtual double train (const vector< vector< double > > &features, const vector< double > &targets)
 estimate the regression parameters a set of training examples
virtual double train (const vector< vector< double > > &features, const vector< double > &targets, const vector< double > &weights)
 estimate the regression parameters a set of weighted training examples
virtual double train (const char *filename)
 estimate the regression parameters from a drwnRegressionDataset file
virtual void getRegressions (const vector< vector< double > > &features, vector< double > &outputTargets) const
 compute the estimated values for a set of feature vector and return the estimates in outputTargets
- Public Member Functions inherited from drwnWriteable
bool write (const char *filename) const
 write object to file (calls save)
bool read (const char *filename)
 read object from file (calls load)
void dump () const
 print object's current state to standard output (for debugging)
- Public Member Functions inherited from drwnCloneable
virtual drwnCloneableclone () const =0
 returns a copy of the class usually implemented as virtual Foo* clone() { return new Foo(*this); }
- Public Member Functions inherited from drwnProperties
unsigned numProperties () const
bool hasProperty (const string &name) const
bool hasProperty (const char *name) const
unsigned findProperty (const string &name) const
unsigned findProperty (const char *name) const
void setProperty (unsigned indx, bool value)
void setProperty (unsigned indx, int value)
void setProperty (unsigned indx, double value)
void setProperty (unsigned indx, const string &value)
void setProperty (unsigned indx, const char *value)
void setProperty (unsigned indx, const Eigen::VectorXd &value)
void setProperty (unsigned indx, const Eigen::MatrixXd &value)
void setProperty (const char *name, bool value)
void setProperty (const char *name, int value)
void setProperty (const char *name, double value)
void setProperty (const char *name, const string &value)
void setProperty (const char *name, const char *value)
void setProperty (const char *name, const Eigen::VectorXd &value)
void setProperty (const char *name, const Eigen::MatrixXd &value)
string getPropertyAsString (unsigned indx) const
drwnPropertyType getPropertyType (unsigned indx) const
bool isReadOnly (unsigned indx) const
const drwnPropertyInterfacegetProperty (unsigned indx) const
const drwnPropertyInterfacegetProperty (const char *name) const
bool getBoolProperty (unsigned indx) const
int getIntProperty (unsigned indx) const
double getDoubleProperty (unsigned indx) const
const string & getStringProperty (unsigned indx) const
const list< string > & getListProperty (unsigned indx) const
int getSelectionProperty (unsigned indx) const
const Eigen::VectorXd & getVectorProperty (unsigned indx) const
const Eigen::MatrixXd & getMatrixProperty (unsigned indx) const
const string & getPropertyName (unsigned indx) const
vector< string > getPropertyNames () const
void readProperties (drwnXMLNode &xml, const char *tag="property")
void writeProperties (drwnXMLNode &xml, const char *tag="property") const
void printProperties (ostream &os) const

Static Public Attributes

static double HUBER_BETA = 1.0e-3
 beta parameter for huber penalty
static double REG_STRENGTH = 1.0e-9
 regularization strength
static int MAX_ITERATIONS = 1000
 maximum training iterations

Protected Member Functions

double objective (const double *x) const
 returns value of objective function at point x
void gradient (const double *x, double *df) const
 populates gradient of objective function at point x
virtual double objectiveAndGradient (const double *x, double *df) const =0
 returns value of objective function and populates gradient df at point x
- Protected Member Functions inherited from drwnProperties
void declareProperty (const string &name, drwnPropertyInterface *optif)
void undeclareProperty (const string &name)
void exposeProperties (drwnProperties *opts, const string &prefix=string(""), bool bSerializable=false)
virtual void propertyChanged (const string &name)
- Protected Member Functions inherited from drwnOptimizer
 drwnOptimizer ()
 default constructor
 drwnOptimizer (unsigned n)
 construct a problem with dimension n
 drwnOptimizer (const drwnOptimizer &o)
 copy constructor
void initialize (unsigned n, const double *x=NULL)
 initialize an optimization problem of size n possibly with feasible starting point x $\mathbb{R}^n$ (or zero)
void initialize (const double *x=NULL)
 initialize an optimization problem at feasible starting point x in $\mathbb{R}^n$ (or zero)
double solve (unsigned maxiter, bool bMonitor=false)
 Solve the optimization problem for up to maxiter iterations to precision set by EPSF, EPSG, and EPSX static variables. Calls monitor function after each iteration if bMonitor is true.
unsigned size () const
 dimension of optimization problem
double operator[] (unsigned i) const
 returns the i-th component of the current solution
double & operator[] (unsigned i)
 returns a reference to the i-th component of the current solution
virtual void monitor (unsigned iter, double objValue)
 callback for each iteration during optimization (if bMonitor is true)

Protected Attributes

VectorXd _theta
 regression weights
int _penalty
 regression penalty option
double _beta
 huber penalty threshold
int _regularizer
 regularization option
double _lambda
 regularization strength
const vector< vector< double > > * _features
const vector< double > * _targets
const vector< double > * _weights
- Protected Attributes inherited from drwnRegression
int _nFeatures
 number of features
bool _bValid
 true if regression parameters are trained or loaded
- Protected Attributes inherited from drwnOptimizer
unsigned _n
 dimension of optimization problem (i.e., $\mathbb{R}^n$)
double * _x
 current feasible solution in $\mathbb{R}^n$
double * _df
 gradient at _x in $\mathbb{R}^n$

Additional Inherited Members

- Static Protected Attributes inherited from drwnOptimizer
static double EPSF = 1.0e-6
 default tolerance on function convergence
static double EPSG = 1.0e-3
 deafult tolerance on gradient convergence
static double EPSX = 1.0e-6
 default tolerance on solution convergence

Detailed Description

Common functionality for drwnLinearRegressor.

The documentation for this class was generated from the following files: