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

The first sentence should be a BRIEF DESCRIPTION of what the class does. More...

#include <MixtureDistribution.h>

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

List of all members.

Public Member Functions

 MixtureDistribution ()
 Default constructor.
virtual real log_density (const Vec &y) const
 Return log of probability density log(p(y | x)).
virtual real survival_fn (const Vec &y) const
 Return survival function: P(Y>y | x).
virtual real cdf (const Vec &y) const
 Return cdf: P(Y<y | x).
virtual void expectation (Vec &mu) const
 Return E[Y | x].
virtual void variance (Mat &cov) const
 Return Var[Y | x].
virtual void generate (Vec &y) const
 Return a pseudo-random sample generated from the conditional distribution, of density p(y | x).
virtual void resetGenerator (long g_seed)
 Generates a pseudo-random sample x from the reversed conditional distribution, of density p(x | y) (and NOT p(y | x)).
virtual bool setPredictorPredictedSizes (int the_predictor_size, int the_predicted_size, bool call_parent=true)
 Set the 'predictor' and 'predicted' sizes for this distribution.
virtual void setPredictor (const Vec &predictor, bool call_parent=true) const
 Set the value for the predictor part of a conditional probability.
virtual void forget ()
 (Re-)initializes the PDistribution in its fresh state (that state may depend on the 'seed' option).
virtual void train ()
 The role of the train method is to bring the learner up to stage == nstages, updating the train_stats collector with training costs measured on-line in the process.
virtual string classname () const
virtual OptionListgetOptionList () const
virtual OptionMapgetOptionMap () const
virtual RemoteMethodMapgetRemoteMethodMap () const
virtual MixtureDistributiondeepCopy (CopiesMap &copies) const
virtual void build ()
 Simply calls inherited::build() then build_().
virtual void makeDeepCopyFromShallowCopy (CopiesMap &copies)
 Transforms a shallow copy into a deep copy.

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 ()

Public Attributes

TVec< PP< PDistribution > > distributions
Vec weights

Static Public Attributes

static StaticInitializer _static_initializer_

Protected Member Functions

void getSizes () const
 Re-obtain the sizes of the predictor and predicted parts from the first distribution in the 'distributions' vector.

Static Protected Member Functions

static void declareOptions (OptionList &ol)
 Declares the class options.

Private Types

typedef PDistribution inherited

Private Member Functions

void build_ ()
 This does the actual building.

Private Attributes

Vec work
 Vector to store temporary data.

Detailed Description

The first sentence should be a BRIEF DESCRIPTION of what the class does.

Place the rest of the class programmer documentation here. Doxygen supports Javadoc-style comments. See http://www.doxygen.org/manual.html

Todo:
Write class to-do's here if there are any.
Deprecated:
Write deprecated stuff here if there is any. Indicate what else should be used instead.

Definition at line 57 of file MixtureDistribution.h.


Member Typedef Documentation

Reimplemented from PLearn::PDistribution.

Definition at line 59 of file MixtureDistribution.h.


Constructor & Destructor Documentation

PLearn::MixtureDistribution::MixtureDistribution ( )

Default constructor.

Definition at line 55 of file MixtureDistribution.cc.

{}

Member Function Documentation

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

Reimplemented from PLearn::PDistribution.

Definition at line 50 of file MixtureDistribution.cc.

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

Reimplemented from PLearn::PDistribution.

Definition at line 50 of file MixtureDistribution.cc.

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

Reimplemented from PLearn::PDistribution.

Definition at line 50 of file MixtureDistribution.cc.

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

Reimplemented from PLearn::PDistribution.

Definition at line 50 of file MixtureDistribution.cc.

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

Reimplemented from PLearn::PDistribution.

Definition at line 50 of file MixtureDistribution.cc.

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

Reimplemented from PLearn::PDistribution.

Definition at line 50 of file MixtureDistribution.cc.

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

Simply calls inherited::build() then build_().

Reimplemented from PLearn::PDistribution.

Definition at line 98 of file MixtureDistribution.cc.

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

This does the actual building.

Reimplemented from PLearn::PDistribution.

Definition at line 107 of file MixtureDistribution.cc.

References PLearn::is_equal(), n, PLCHECK_MSG, and PLearn::sum().

{
    if (distributions.isEmpty())
        return;
    if (weights.isEmpty()) {
        int n = distributions.length();
        weights.resize(n);
        weights.fill(1 / real(n));
    }
    PLCHECK_MSG(weights.length() == distributions.length() &&
                 is_equal(PLearn::sum(weights), 1),
                 "There must be one weight for each distribution, and the "
                 "weights must sum to 1");
    getSizes();
}

Here is the call graph for this function:

real PLearn::MixtureDistribution::cdf ( const Vec y) const [virtual]

Return cdf: P(Y<y | x).

Reimplemented from PLearn::PDistribution.

Definition at line 126 of file MixtureDistribution.cc.

References PLERROR.

{
    PLERROR("cdf not implemented for MixtureDistribution"); return 0;
}
string PLearn::MixtureDistribution::classname ( ) const [virtual]

Reimplemented from PLearn::PDistribution.

Definition at line 50 of file MixtureDistribution.cc.

void PLearn::MixtureDistribution::declareOptions ( OptionList ol) [static, protected]

Declares the class options.

Reimplemented from PLearn::PDistribution.

Definition at line 61 of file MixtureDistribution.cc.

References PLearn::OptionBase::buildoption, PLearn::declareOption(), distributions, PLearn::OptionBase::nosave, PLearn::PDistribution::predicted_size, PLearn::PDistribution::predictor_part, PLearn::PDistribution::predictor_size, PLearn::redeclareOption(), and weights.

{
    declareOption(ol, "distributions", &MixtureDistribution::distributions,
                  OptionBase::buildoption,
        "Underlying distributions being mixed.");

    declareOption(ol, "weights", &MixtureDistribution::weights,
                  OptionBase::buildoption,
        "Weights of the distributions (must sum to 1). If left empty, then\n"
        "each distribution will be given a weight 1/number_of_distributions.");

    // Now call the parent class' declareOptions().
    inherited::declareOptions(ol);

    // Hide unused options.

    redeclareOption(ol, "predicted_size",
                    &MixtureDistribution::predicted_size,
                    OptionBase::nosave,
        "Unused");

    redeclareOption(ol, "predictor_part",
                    &MixtureDistribution::predictor_part,
                    OptionBase::nosave,
        "Unused");

    redeclareOption(ol, "predictor_size",
                    &MixtureDistribution::predictor_size,
                    OptionBase::nosave,
        "Unused");


}

Here is the call graph for this function:

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

Reimplemented from PLearn::PDistribution.

Definition at line 175 of file MixtureDistribution.h.

:
    //#####  Protected Options  ###############################################
MixtureDistribution * PLearn::MixtureDistribution::deepCopy ( CopiesMap copies) const [virtual]

Reimplemented from PLearn::PDistribution.

Definition at line 50 of file MixtureDistribution.cc.

void PLearn::MixtureDistribution::expectation ( Vec mu) const [virtual]

Return E[Y | x].

Reimplemented from PLearn::PDistribution.

Definition at line 134 of file MixtureDistribution.cc.

References PLearn::TVec< T >::fill(), i, PLearn::multiplyAcc(), PLASSERT, and PLearn::TVec< T >::resize().

{
    PLASSERT( !distributions.isEmpty() );
    mu.resize(distributions[0]->getNPredicted());
    mu.fill(0);
    for (int i = 0; i < distributions.length(); i++) {
        distributions[i]->expectation(work);
        multiplyAcc(mu, work, weights[i]);
    }
}

Here is the call graph for this function:

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

(Re-)initializes the PDistribution 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!). ### You may remove this method if your distribution does not ### implement it.

Reimplemented from PLearn::PDistribution.

Definition at line 148 of file MixtureDistribution.cc.

References i.

{
    for (int i = 0; i < distributions.length(); i++)
        distributions[i]->forget();
    inherited::forget();
    getSizes();
}
void PLearn::MixtureDistribution::generate ( Vec y) const [virtual]

Return a pseudo-random sample generated from the conditional distribution, of density p(y | x).

Reimplemented from PLearn::PDistribution.

Definition at line 159 of file MixtureDistribution.cc.

References j.

{
    int j = random_gen->multinomial_sample(weights);
    distributions[j]->generate(y);
}
OptionList & PLearn::MixtureDistribution::getOptionList ( ) const [virtual]

Reimplemented from PLearn::PDistribution.

Definition at line 50 of file MixtureDistribution.cc.

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

Reimplemented from PLearn::PDistribution.

Definition at line 50 of file MixtureDistribution.cc.

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

Reimplemented from PLearn::PDistribution.

Definition at line 50 of file MixtureDistribution.cc.

void PLearn::MixtureDistribution::getSizes ( ) const [protected]

Re-obtain the sizes of the predictor and predicted parts from the first distribution in the 'distributions' vector.

This method is used to re-obtain sizes after things may have changed (e.g. after a build(), forget() or train()).

Definition at line 168 of file MixtureDistribution.cc.

References PLASSERT.

                                         {
    PLASSERT( !distributions.isEmpty() );
    n_predicted = distributions[0]->getNPredicted();
    n_predictor = distributions[0]->getNPredictor();
}
real PLearn::MixtureDistribution::log_density ( const Vec y) const [virtual]

Return log of probability density log(p(y | x)).

Reimplemented from PLearn::PDistribution.

Definition at line 177 of file MixtureDistribution.cc.

References i, PLearn::logadd(), n, and pl_log.

{
    int n = distributions.length();
    work.resize(n);
    for (int i = 0; i < n; i++) {
        work[i] = distributions[i]->log_density(y) + pl_log(weights[i]);
    }
    return logadd(work);
}

Here is the call graph for this function:

void PLearn::MixtureDistribution::makeDeepCopyFromShallowCopy ( CopiesMap copies) [virtual]

Transforms a shallow copy into a deep copy.

Reimplemented from PLearn::PDistribution.

Definition at line 190 of file MixtureDistribution.cc.

References PLERROR.

{
    inherited::makeDeepCopyFromShallowCopy(copies);

    // ### Call deepCopyField on all "pointer-like" fields
    // ### that you wish to be deepCopied rather than
    // ### shallow-copied.
    // ### ex:
    // deepCopyField(trainvec, copies);

    // ### Remove this line when you have fully implemented this method.
    PLERROR("MixtureDistribution::makeDeepCopyFromShallowCopy not fully (correctly) implemented yet!");
}
void PLearn::MixtureDistribution::resetGenerator ( long  g_seed) [virtual]

Generates a pseudo-random sample x from the reversed conditional distribution, of density p(x | y) (and NOT p(y | x)).

i.e., generates a "predictor" part given a "predicted" part, regardless of any previously set predictor. Reset the random number generator used by generate() using the given seed.

Reimplemented from PLearn::PDistribution.

Definition at line 207 of file MixtureDistribution.cc.

References i.

{
    for (int i = 0; i < distributions.length(); i++)
        distributions[i]->resetGenerator(g_seed);
    inherited::resetGenerator(g_seed);
}
void PLearn::MixtureDistribution::setPredictor ( const Vec predictor,
bool  call_parent = true 
) const [virtual]

Set the value for the predictor part of a conditional probability.

Reimplemented from PLearn::PDistribution.

Definition at line 217 of file MixtureDistribution.cc.

References i.

{
    if (call_parent)
        inherited::setPredictor(predictor, true);
    for (int i = 0; i < distributions.length(); i++)
        distributions[i]->setPredictor(predictor, call_parent);
    getSizes();
}
bool PLearn::MixtureDistribution::setPredictorPredictedSizes ( int  the_predictor_size,
int  the_predicted_size,
bool  call_parent = true 
) [virtual]

Set the 'predictor' and 'predicted' sizes for this distribution.

Reimplemented from PLearn::PDistribution.

Definition at line 229 of file MixtureDistribution.cc.

References i.

{
    bool sizes_have_changed = false;
    if (call_parent)
        sizes_have_changed = inherited::setPredictorPredictedSizes(
                the_predictor_size, the_predicted_size, true);
    for (int i = 0; i < distributions.length(); i++)
        distributions[i]->setPredictorPredictedSizes(the_predictor_size,
                                                     the_predicted_size,
                                                     call_parent);
    getSizes();
    // Returned value.
    return sizes_have_changed;
}
real PLearn::MixtureDistribution::survival_fn ( const Vec y) const [virtual]

Return survival function: P(Y>y | x).

Reimplemented from PLearn::PDistribution.

Definition at line 249 of file MixtureDistribution.cc.

References PLERROR.

{
    PLERROR("survival_fn not implemented for MixtureDistribution"); return 0;
}
void PLearn::MixtureDistribution::train ( ) [virtual]

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

Reimplemented from PLearn::PDistribution.

Definition at line 257 of file MixtureDistribution.cc.

References i, and PLCHECK.

{
    // This generic PLearner method does a number of standard stuff useful for
    // (almost) any learner, and return 'false' if no training should take
    // place. See PLearner.h for more details.
    if (!initTrain())
        return;

    PLCHECK( nstages == 1 && stage == 0 );
    for (int i = 0; i < distributions.length(); i++)
        distributions[i]->train();
    stage = 1;
    getSizes();
}
void PLearn::MixtureDistribution::variance ( Mat cov) const [virtual]

Return Var[Y | x].

Reimplemented from PLearn::PDistribution.

Definition at line 275 of file MixtureDistribution.cc.

References PLERROR.

{
    PLERROR("variance not implemented for MixtureDistribution");
}

Member Data Documentation

Reimplemented from PLearn::PDistribution.

Definition at line 175 of file MixtureDistribution.h.

Definition at line 64 of file MixtureDistribution.h.

Referenced by declareOptions().

Definition at line 65 of file MixtureDistribution.h.

Referenced by declareOptions().

Vector to store temporary data.

Definition at line 211 of file MixtureDistribution.h.


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