PLearn 0.1
Public Member Functions | Static Public Member Functions | Static Public Attributes | Private Types | Private Member Functions | Private Attributes
PLearn::BaseRegressorWrapper Class Reference

#include <BaseRegressorWrapper.h>

Inheritance diagram for PLearn::BaseRegressorWrapper:
Inheritance graph
[legend]
Collaboration diagram for PLearn::BaseRegressorWrapper:
Collaboration graph
[legend]

List of all members.

Public Member Functions

 BaseRegressorWrapper ()
virtual ~BaseRegressorWrapper ()
virtual string classname () const
virtual OptionListgetOptionList () const
virtual OptionMapgetOptionMap () const
virtual RemoteMethodMapgetRemoteMethodMap () const
virtual BaseRegressorWrapperdeepCopy (CopiesMap &copies) const
virtual void makeDeepCopyFromShallowCopy (CopiesMap &copies)
 Transforms a shallow copy into a deep copy.
virtual void build ()
 Finish building the object; just call inherited::build followed by build_()
virtual void train ()
 *** SUBCLASS WRITING: ***
virtual void forget ()
 *** SUBCLASS WRITING: ***
virtual int outputsize () const
 SUBCLASS WRITING: override this so that it returns the size of this learner's output, as a function of its inputsize(), targetsize() and set options.
virtual TVec< string > getTrainCostNames () const
 *** SUBCLASS WRITING: ***
virtual TVec< string > getTestCostNames () const
 *** SUBCLASS WRITING: ***
virtual void computeOutput (const Vec &input, Vec &output) const
 *** SUBCLASS WRITING: ***
virtual void computeOutputAndCosts (const Vec &input, const Vec &target, Vec &output, Vec &costs) const
 Default calls computeOutput and computeCostsFromOutputs.
virtual void computeCostsFromOutputs (const Vec &input, const Vec &output, const Vec &target, Vec &costs) const
 *** SUBCLASS WRITING: ***
void setSortedTrainSet (PP< RegressionTreeRegisters > the_sorted_train_set)

Static Public Member Functions

static string _classname_ ()
static OptionList_getOptionList_ ()
static RemoteMethodMap_getRemoteMethodMap_ ()
static Object_new_instance_for_typemap_ ()
static bool _isa_ (const Object *o)
static void _static_initialize_ ()
static const PPathdeclaringFile ()
static void declareOptions (OptionList &ol)
 Declares this class' options.

Static Public Attributes

static StaticInitializer _static_initializer_

Private Types

typedef PLearner inherited

Private Member Functions

void build_ ()
 **** SUBCLASS WRITING: ****
void verbose (string msg, int level)

Private Attributes

real loss_function_weight
int mean_imputation
int regression_tree
int use_confidence_function
int use_base_regressor_confidence
PP< PLearnerbase_regressor_template
PP< RegressionTreetree_regressor_template
PP< BaseRegressorConfidenceconfidence_function_template
PP< RegressionTreeRegisterssorted_train_set
PP< PLearnerbase_regressor
PP< RegressionTreetree_regressor
PP< BaseRegressorConfidenceconfidence_function
VMat base_regressor_train_set
TVec< realvariable_means

Detailed Description

Definition at line 55 of file BaseRegressorWrapper.h.


Member Typedef Documentation

Reimplemented from PLearn::PLearner.

Definition at line 57 of file BaseRegressorWrapper.h.


Constructor & Destructor Documentation

PLearn::BaseRegressorWrapper::BaseRegressorWrapper ( )
PLearn::BaseRegressorWrapper::~BaseRegressorWrapper ( ) [virtual]

Definition at line 71 of file BaseRegressorWrapper.cc.

{
}

Member Function Documentation

string PLearn::BaseRegressorWrapper::_classname_ ( ) [static]

Reimplemented from PLearn::PLearner.

Definition at line 60 of file BaseRegressorWrapper.cc.

OptionList & PLearn::BaseRegressorWrapper::_getOptionList_ ( ) [static]

Reimplemented from PLearn::PLearner.

Definition at line 60 of file BaseRegressorWrapper.cc.

RemoteMethodMap & PLearn::BaseRegressorWrapper::_getRemoteMethodMap_ ( ) [static]

Reimplemented from PLearn::PLearner.

Definition at line 60 of file BaseRegressorWrapper.cc.

bool PLearn::BaseRegressorWrapper::_isa_ ( const Object o) [static]

Reimplemented from PLearn::PLearner.

Definition at line 60 of file BaseRegressorWrapper.cc.

Object * PLearn::BaseRegressorWrapper::_new_instance_for_typemap_ ( ) [static]

Reimplemented from PLearn::Object.

Definition at line 60 of file BaseRegressorWrapper.cc.

StaticInitializer BaseRegressorWrapper::_static_initializer_ & PLearn::BaseRegressorWrapper::_static_initialize_ ( ) [static]

Reimplemented from PLearn::PLearner.

Definition at line 60 of file BaseRegressorWrapper.cc.

void PLearn::BaseRegressorWrapper::build ( ) [virtual]

Finish building the object; just call inherited::build followed by build_()

Reimplemented from PLearn::PLearner.

Definition at line 128 of file BaseRegressorWrapper.cc.

References PLearn::PLearner::build(), and build_().

Here is the call graph for this function:

void PLearn::BaseRegressorWrapper::build_ ( ) [private]

**** SUBCLASS WRITING: ****

This method should finish building of the object, according to set 'options', in *any* situation.

Typical situations include:

  • Initial building of an object from a few user-specified options
  • Building of a "reloaded" object: i.e. from the complete set of all serialised options.
  • Updating or "re-building" of an object after a few "tuning" options (such as hyper-parameters) have been modified.

You can assume that the parent class' build_() has already been called.

A typical build method will want to know the inputsize(), targetsize() and outputsize(), and may also want to check whether train_set->hasWeights(). All these methods require a train_set to be set, so the first thing you may want to do, is check if(train_set), before doing any heavy building...

Note: build() is always called by setTrainingSet.

Reimplemented from PLearn::PLearner.

Definition at line 134 of file BaseRegressorWrapper.cc.

References base_regressor, base_regressor_template, base_regressor_train_set, confidence_function, confidence_function_template, PLearn::deepCopy(), loss_function_weight, mean_imputation, regression_tree, sorted_train_set, PLearn::tostring(), PLearn::PLearner::train_set, tree_regressor, tree_regressor_template, use_confidence_function, and variable_means.

Referenced by build().

{
    if (train_set)
    {
        if (mean_imputation > 0)
        {
            PP<MeanImputationVMatrix> new_train_set = new MeanImputationVMatrix(train_set, 0.0);
            variable_means = new_train_set->getMeanVector();
            base_regressor_train_set = VMat(new_train_set);
        }
        else
        {
            base_regressor_train_set = train_set;
        }
        if (regression_tree > 0)
        {
            tree_regressor = ::PLearn::deepCopy(tree_regressor_template);
            tree_regressor->setTrainingSet(VMat(sorted_train_set));
            tree_regressor->setOption("loss_function_weight", tostring(loss_function_weight));
            base_regressor = tree_regressor;
        }
        else
        {
            base_regressor = ::PLearn::deepCopy(base_regressor_template);
        }
        base_regressor->setTrainingSet(base_regressor_train_set, true);
        base_regressor->setTrainStatsCollector(new VecStatsCollector);
        if (use_confidence_function > 0)
        {
            confidence_function = ::PLearn::deepCopy(confidence_function_template);
            confidence_function->setTrainingSet(base_regressor_train_set, true);
            confidence_function->setTrainStatsCollector(new VecStatsCollector);
            confidence_function->train();
        }
    }
}

Here is the call graph for this function:

Here is the caller graph for this function:

string PLearn::BaseRegressorWrapper::classname ( ) const [virtual]

Reimplemented from PLearn::Object.

Definition at line 60 of file BaseRegressorWrapper.cc.

void PLearn::BaseRegressorWrapper::computeCostsFromOutputs ( const Vec input,
const Vec output,
const Vec target,
Vec costs 
) const [virtual]

*** SUBCLASS WRITING: ***

This should be defined in subclasses to compute the weighted costs from already computed output. The costs should correspond to the cost names returned by getTestCostNames().

NOTE: In exotic cases, the cost may also depend on some info in the input, that's why the method also gets so see it.

Implements PLearn::PLearner.

Definition at line 265 of file BaseRegressorWrapper.cc.

References loss_function_weight, and PLearn::pow().

Referenced by computeOutputAndCosts().

{
    costsv[0] = pow((outputv[0] - targetv[0]), 2);
    costsv[1] = outputv[1];
    costsv[2] = 1.0 - (2.0 * loss_function_weight * costsv[0]);
}

Here is the call graph for this function:

Here is the caller graph for this function:

void PLearn::BaseRegressorWrapper::computeOutput ( const Vec input,
Vec output 
) const [virtual]

*** SUBCLASS WRITING: ***

This should be defined in subclasses to compute the output from the input.

Reimplemented from PLearn::PLearner.

Definition at line 222 of file BaseRegressorWrapper.cc.

References base_regressor, confidence_function, PLearn::is_missing(), mean_imputation, PLearn::TVec< T >::resize(), PLearn::PLearner::train_set, use_base_regressor_confidence, use_confidence_function, and variable_means.

Referenced by computeOutputAndCosts().

{
    Vec base_regressor_inputv;
    Vec base_regressor_outputv;
    Vec confidence_outputv;
    base_regressor_inputv.resize(train_set->inputsize());
    base_regressor_outputv.resize(base_regressor->outputsize());
    for (int variable = 0; variable < train_set->inputsize(); variable++)
    {
        base_regressor_inputv[variable] = inputv[variable];
        if (mean_imputation > 0 && is_missing(inputv[variable]) ) base_regressor_inputv[variable] = variable_means[variable];
    }
    base_regressor->computeOutput(base_regressor_inputv, base_regressor_outputv);
    outputv[0] = base_regressor_outputv[0];
//  cout << "base regressor output: " << outputv[0];
    if (use_base_regressor_confidence > 0)
    {
        outputv[1] = base_regressor_outputv[1];
    }
    else
    {
        if (use_confidence_function > 0)
        {
            confidence_outputv.resize(confidence_function->outputsize());
            confidence_outputv[0] = base_regressor_outputv[0];
            confidence_function->computeOutput(base_regressor_inputv, confidence_outputv);
            outputv[1] = confidence_outputv[1];
        }
        else
        {
            outputv[1] = 1.0;
        }
    }
//  cout << "confidence output: " << confidence_outputv[0];
//  cout << " confidence: " << outputv[1] << endl;
}

Here is the call graph for this function:

Here is the caller graph for this function:

void PLearn::BaseRegressorWrapper::computeOutputAndCosts ( const Vec input,
const Vec target,
Vec output,
Vec costs 
) const [virtual]

Default calls computeOutput and computeCostsFromOutputs.

You may override this if you have a more efficient way to compute both output and weighted costs at the same time.

Reimplemented from PLearn::PLearner.

Definition at line 259 of file BaseRegressorWrapper.cc.

References computeCostsFromOutputs(), and computeOutput().

{
    computeOutput(inputv, outputv);
    computeCostsFromOutputs(inputv, outputv, targetv, costsv);
}

Here is the call graph for this function:

void PLearn::BaseRegressorWrapper::declareOptions ( OptionList ol) [static]

Declares this class' options.

Reimplemented from PLearn::PLearner.

Definition at line 75 of file BaseRegressorWrapper.cc.

References base_regressor, base_regressor_template, base_regressor_train_set, PLearn::OptionBase::buildoption, confidence_function, confidence_function_template, PLearn::declareOption(), PLearn::PLearner::declareOptions(), PLearn::OptionBase::learntoption, loss_function_weight, mean_imputation, regression_tree, sorted_train_set, tree_regressor_template, use_base_regressor_confidence, use_confidence_function, and variable_means.

{ 
    declareOption(ol, "loss_function_weight", &BaseRegressorWrapper::loss_function_weight, OptionBase::buildoption,
                  "The hyper parameter to balance the error and the confidence factor.\n");
    declareOption(ol, "mean_imputation", &BaseRegressorWrapper::mean_imputation, OptionBase::buildoption,
                  "If set to 1, the algorithm will compute the mean vector of the input variables from the training set,\n"
                  "and replace missing values with the computed means.\n");
    declareOption(ol, "regression_tree", &BaseRegressorWrapper::regression_tree, OptionBase::buildoption,
                  "If set to 1, the tree_regressor_template is used instead of the base_regressor_template.\n"
                  "It permits to sort the train set only once for all boosting iterations.\n");       
    declareOption(ol, "use_confidence_function", &BaseRegressorWrapper::use_confidence_function, OptionBase::buildoption,
                  "If set to 1, the confidence_function_template is used to build a confidence estimates on the base regressor prediction.\n"
                  "Otherwise, if the confidence from the base regressor is not used, the confidence is always set to 1.0.\n");
    declareOption(ol, "use_base_regressor_confidence", &BaseRegressorWrapper::use_base_regressor_confidence, OptionBase::buildoption,
                  "If set to 1, the confidence is taken from the second output of the base regressor.\n");
    declareOption(ol, "base_regressor_template", &BaseRegressorWrapper::base_regressor_template, OptionBase::buildoption,
                  "The template for the base regressor to be boosted.\n"); 
    declareOption(ol, "tree_regressor_template", &BaseRegressorWrapper::tree_regressor_template, OptionBase::buildoption,
                  "The template for a RegressionTree base regressor when the regression_tree option is set to 1.\n");  
    declareOption(ol, "confidence_function_template", &BaseRegressorWrapper::confidence_function_template, OptionBase::buildoption,
                  "The template for the confidence function to be learnt from the train set.\n"); 
    declareOption(ol, "sorted_train_set", &BaseRegressorWrapper::sorted_train_set, OptionBase::buildoption,
                  "A sorted train set when using a tree as a base regressor\n");
      
    declareOption(ol, "base_regressor", &BaseRegressorWrapper::base_regressor, OptionBase::learntoption,
                  "The base regressor built from the template.\n");
    declareOption(ol, "confidence_function", &BaseRegressorWrapper::confidence_function, OptionBase::learntoption,
                  "The confidence function learnt from the train set.\n");
    declareOption(ol, "base_regressor_train_set", &BaseRegressorWrapper::base_regressor_train_set, OptionBase::learntoption,
                  "The VMat train set prepared for the base regressor.\n"
                  "It apllies the chosen missing value management strategies.\n");
    declareOption(ol, "variable_means", &BaseRegressorWrapper::variable_means, OptionBase::learntoption,
                  "The vector with the computed means on all input dimension to perform mean imputation if applicable.\n");
    inherited::declareOptions(ol);
}

Here is the call graph for this function:

static const PPath& PLearn::BaseRegressorWrapper::declaringFile ( ) [inline, static]

Reimplemented from PLearn::PLearner.

Definition at line 96 of file BaseRegressorWrapper.h.

:
    void         build_();
BaseRegressorWrapper * PLearn::BaseRegressorWrapper::deepCopy ( CopiesMap copies) const [virtual]

Reimplemented from PLearn::PLearner.

Definition at line 60 of file BaseRegressorWrapper.cc.

void PLearn::BaseRegressorWrapper::forget ( ) [virtual]

*** SUBCLASS WRITING: ***

(Re-)initializes the PLearner in its fresh state (that state may depend on the 'seed' option) and sets 'stage' back to 0 (this is the stage of a fresh learner!)

A typical forget() method should do the following:

  • initialize the learner's parameters, using this random generator
  • stage = 0;

This method is typically called by the build_() method, after it has finished setting up the parameters, and if it deemed useful to set or reset the learner in its fresh state. (remember build may be called after modifying options that do not necessarily require the learner to restart from a fresh state...) forget is also called by the setTrainingSet method, after calling build(), so it will generally be called TWICE during setTrainingSet!

Reimplemented from PLearn::PLearner.

Definition at line 199 of file BaseRegressorWrapper.cc.

{
}
OptionList & PLearn::BaseRegressorWrapper::getOptionList ( ) const [virtual]

Reimplemented from PLearn::Object.

Definition at line 60 of file BaseRegressorWrapper.cc.

OptionMap & PLearn::BaseRegressorWrapper::getOptionMap ( ) const [virtual]

Reimplemented from PLearn::Object.

Definition at line 60 of file BaseRegressorWrapper.cc.

RemoteMethodMap & PLearn::BaseRegressorWrapper::getRemoteMethodMap ( ) const [virtual]

Reimplemented from PLearn::Object.

Definition at line 60 of file BaseRegressorWrapper.cc.

TVec< string > PLearn::BaseRegressorWrapper::getTestCostNames ( ) const [virtual]

*** SUBCLASS WRITING: ***

This should return the names of the costs computed by computeCostsFromOutputs.

Implements PLearn::PLearner.

Definition at line 217 of file BaseRegressorWrapper.cc.

References getTrainCostNames().

{ 
    return getTrainCostNames();
}

Here is the call graph for this function:

TVec< string > PLearn::BaseRegressorWrapper::getTrainCostNames ( ) const [virtual]

*** SUBCLASS WRITING: ***

This should return the names of the objective costs that the train method computes and for which it updates the VecStatsCollector train_stats.

Implements PLearn::PLearner.

Definition at line 208 of file BaseRegressorWrapper.cc.

Referenced by getTestCostNames().

{
    TVec<string> return_msg(3);
    return_msg[0] = "mse";
    return_msg[1] = "base confidence";
    return_msg[2] = "base reward";
    return return_msg;
}

Here is the caller graph for this function:

void PLearn::BaseRegressorWrapper::makeDeepCopyFromShallowCopy ( CopiesMap copies) [virtual]
int PLearn::BaseRegressorWrapper::outputsize ( ) const [virtual]

SUBCLASS WRITING: override this so that it returns the size of this learner's output, as a function of its inputsize(), targetsize() and set options.

Implements PLearn::PLearner.

Definition at line 203 of file BaseRegressorWrapper.cc.

{
    return 2;
}
void PLearn::BaseRegressorWrapper::setSortedTrainSet ( PP< RegressionTreeRegisters the_sorted_train_set)

Definition at line 272 of file BaseRegressorWrapper.cc.

References sorted_train_set.

{
    sorted_train_set = the_sorted_train_set;
}
void PLearn::BaseRegressorWrapper::train ( ) [virtual]

*** SUBCLASS WRITING: ***

The role of the train method is to bring the learner up to stage==nstages, updating the stats with training costs measured on-line in the process.

TYPICAL CODE:

  static Vec input;  // static so we don't reallocate/deallocate memory each time...
  static Vec target; // (but be careful that static means shared!)
  input.resize(inputsize());    // the train_set's inputsize()
  target.resize(targetsize());  // the train_set's targetsize()
  real weight;
  
  if(!train_stats)   // make a default stats collector, in case there's none
      train_stats = new VecStatsCollector();
  
  if(nstages<stage)  // asking to revert to a previous stage!
      forget();      // reset the learner to stage=0
  
  while(stage<nstages)
  {
      // clear statistics of previous epoch
      train_stats->forget(); 
            
      //... train for 1 stage, and update train_stats,
      // using train_set->getSample(input, target, weight);
      // and train_stats->update(train_costs)
          
      ++stage;
      train_stats->finalize(); // finalize statistics for this epoch
  }

Implements PLearn::PLearner.

Definition at line 171 of file BaseRegressorWrapper.cc.

References base_regressor.

{
    base_regressor->train();
    /*
      cout << "testing the confidence function" << endl;
      Vec train_input;
      Vec train_target;
      Vec train_output;
      real train_weight;
      train_input->resize(train_set->inputsize());
      train_target->resize(train_set->targetsize());
      train_output->resize(2);
      for (int row = 0; row < 25; row++)  
      {
      train_set->getExample(row, train_input, train_target, train_weight);
      cout << "row: " << row << " target: " << train_target[0];
      computeOutput(train_input, train_output);
      }
      PLERROR("We are done for now");
    */
}
void PLearn::BaseRegressorWrapper::verbose ( string  msg,
int  level 
) [private]

Definition at line 193 of file BaseRegressorWrapper.cc.

References PLearn::endl(), and PLearn::PLearner::verbosity.

{
    if (verbosity >= the_level)
        cout << the_msg << endl;
}

Here is the call graph for this function:


Member Data Documentation

Reimplemented from PLearn::PLearner.

Definition at line 96 of file BaseRegressorWrapper.h.

Definition at line 70 of file BaseRegressorWrapper.h.

Referenced by build_(), declareOptions(), and makeDeepCopyFromShallowCopy().

Definition at line 83 of file BaseRegressorWrapper.h.

Referenced by build_(), declareOptions(), and makeDeepCopyFromShallowCopy().

Definition at line 72 of file BaseRegressorWrapper.h.

Referenced by build_(), declareOptions(), and makeDeepCopyFromShallowCopy().

Definition at line 67 of file BaseRegressorWrapper.h.

Referenced by build_(), declareOptions(), and makeDeepCopyFromShallowCopy().

Definition at line 81 of file BaseRegressorWrapper.h.

Referenced by build_().

Definition at line 71 of file BaseRegressorWrapper.h.

Referenced by build_(), declareOptions(), and makeDeepCopyFromShallowCopy().

Definition at line 84 of file BaseRegressorWrapper.h.

Referenced by build_(), computeOutput(), and declareOptions().


The documentation for this class was generated from the following files:
 All Classes Namespaces Files Functions Variables Typedefs Enumerations Enumerator Friends Defines