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

Squared difference (and derivatives thereof) between the target and input. More...

#include <SquaredErrModule.h>

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

List of all members.

Public Member Functions

 SquaredErrModule ()
 Default constructor.
virtual void fprop (const Vec &input, Vec &output) const
 given the input, compute the output (possibly resize it appropriately) SOON TO BE DEPRECATED, USE fprop(const TVec<Mat*>& ports_value)
virtual void bpropUpdate (const Vec &input, const Vec &output, const Vec &output_gradient)
 SOON TO BE DEPRECATED, USE bpropAccUpdate(const TVec<Mat*>& ports_value, const TVec<Mat*>& ports_gradient) Adapt based on the output gradient: this method should only be called just after a corresponding fprop; it should be called with the same arguments as fprop for the first two arguments (and output should not have been modified since then).
virtual void bpropUpdate (const Vec &input, const Vec &output, Vec &input_gradient, const Vec &output_gradient)
virtual void bbpropUpdate (const Vec &input, const Vec &output, const Vec &output_gradient, const Vec &output_diag_hessian)
 Similar to bpropUpdate, but adapt based also on the estimation of the diagonal of the Hessian matrix, and propagates this back.
virtual void bbpropUpdate (const Vec &input, const Vec &output, Vec &input_gradient, const Vec &output_gradient, Vec &input_diag_hessian, const Vec &output_diag_hessian)
virtual void forget ()
 reset the parameters to the state they would be BEFORE starting training.
virtual string classname () const
virtual OptionListgetOptionList () const
virtual OptionMapgetOptionMap () const
virtual RemoteMethodMapgetRemoteMethodMap () const
virtual SquaredErrModuledeepCopy (CopiesMap &copies) const
virtual void build ()
 Post-constructor.
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 ()

Static Public Attributes

static StaticInitializer _static_initializer_

Static Protected Member Functions

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

Protected Attributes

int target_size

Private Types

typedef OnlineLearningModule inherited

Private Member Functions

void build_ ()
 This does the actual building.

Detailed Description

Squared difference (and derivatives thereof) between the target and input.

Deprecated:
: Use ../SquaredErrorCostModule instead

Definition at line 58 of file SquaredErrModule.h.


Member Typedef Documentation

Reimplemented from PLearn::OnlineLearningModule.

Definition at line 60 of file SquaredErrModule.h.


Constructor & Destructor Documentation

PLearn::SquaredErrModule::SquaredErrModule ( )

Default constructor.

Definition at line 55 of file SquaredErrModule.cc.

References PLearn::OnlineLearningModule::output_size.

                                   :
    /* ### Initialize all fields to their default value */
    target_size( 0 )
{
    output_size = 1;
}

Member Function Documentation

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

Reimplemented from PLearn::OnlineLearningModule.

Definition at line 53 of file SquaredErrModule.cc.

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

Reimplemented from PLearn::OnlineLearningModule.

Definition at line 53 of file SquaredErrModule.cc.

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

Reimplemented from PLearn::OnlineLearningModule.

Definition at line 53 of file SquaredErrModule.cc.

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

Reimplemented from PLearn::OnlineLearningModule.

Definition at line 53 of file SquaredErrModule.cc.

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

Reimplemented from PLearn::Object.

Definition at line 53 of file SquaredErrModule.cc.

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

Reimplemented from PLearn::OnlineLearningModule.

Definition at line 53 of file SquaredErrModule.cc.

void PLearn::SquaredErrModule::bbpropUpdate ( const Vec input,
const Vec output,
Vec input_gradient,
const Vec output_gradient,
Vec input_diag_hessian,
const Vec output_diag_hessian 
) [virtual]

Definition at line 180 of file SquaredErrModule.cc.

References bpropUpdate(), PLearn::TVec< T >::clear(), PLearn::OnlineLearningModule::estimate_simpler_diag_hessian, i, PLearn::OnlineLearningModule::input_size, PLearn::OnlineLearningModule::output_size, PLERROR, PLWARNING, PLearn::TVec< T >::resize(), PLearn::TVec< T >::size(), PLearn::TVec< T >::subVec(), and target_size.

{
    int odh_size = output_diag_hessian.size();
    bool is_final_cost = false; // if yes, output_diag_hessian is 0

    // size check
    // others size checks will be done in bpropUpdate()
    if( odh_size == 0 )
    {
        PLWARNING("SquaredErrModule::bbpropUpdate: you are not providing"
                  " output_diag_hessian.\n"
                  "Assuming this is the final cost,"
                  " and output_diag_hessian=0.\n");
        is_final_cost = true;
    }
    else if( odh_size != output_size )
    {
        PLERROR("SquaredErrModule::bbpropUpdate:"
                " 'output_diag_hessian.size()'\n"
                " should be equal to 'output_size' (%i != %i)\n",
                odh_size, output_size);
    }

    bpropUpdate( input, output, input_gradient, output_gradient );

    Vec input_ = input.subVec( 0, input_size );
    Vec target = input.subVec( input_size, target_size );
    input_diag_hessian.resize( input_size );

    // computation of term dC/dy d²y/dx²,
    // skipped if estimate_simpler_diag_hessian, unless it is final cost
    if( estimate_simpler_diag_hessian && !is_final_cost )
    {
        input_diag_hessian.clear();
    }
    else
    {
        Vec idh( input_size, 2 );
        input_diag_hessian << idh;

        if( !is_final_cost )
            input_diag_hessian *= output_gradient[0];
    }

    // computation of term d²C/dy² (dy/dx)²,
    // skipped if it is final cost, because then d²C/dy² == d²C/dC² == 0
    if( !is_final_cost )
    {
        for( int i=0 ; i<input_size ; i++ )
        {
            real fprime_i = 2*(input_[i] - target[i]);
            input_diag_hessian[i] += (fprime_i*fprime_i)
                                       * output_diag_hessian[0];
        }
    }

}

Here is the call graph for this function:

void PLearn::SquaredErrModule::bbpropUpdate ( const Vec input,
const Vec output,
const Vec output_gradient,
const Vec output_diag_hessian 
) [virtual]

Similar to bpropUpdate, but adapt based also on the estimation of the diagonal of the Hessian matrix, and propagates this back.

If these methods are defined, you can use them INSTEAD of bpropUpdate(...) THE DEFAULT IMPLEMENTATION PROVIDED HERE JUST CALLS bbpropUpdate(input, output, input_gradient, output_gradient, in_hess, out_hess) AND IGNORES INPUT HESSIAN AND INPUT GRADIENT

Reimplemented from PLearn::OnlineLearningModule.

Definition at line 163 of file SquaredErrModule.cc.

References bpropUpdate(), PLearn::OnlineLearningModule::output_size, PLWARNING, and PLearn::TVec< T >::size().

{
    int odh_size = output_diag_hessian.size();
    if( odh_size != output_size )
    {
        PLWARNING("SquaredErrModule::bbpropUpdate:"
                  " 'output_diag_hessian.size()'\n"
                  " should be equal to 'output_size' (%i != %i)\n",
                  odh_size, output_size);
    }

    bpropUpdate( input, output, output_gradient );
}

Here is the call graph for this function:

void PLearn::SquaredErrModule::bpropUpdate ( const Vec input,
const Vec output,
const Vec output_gradient 
) [virtual]

SOON TO BE DEPRECATED, USE bpropAccUpdate(const TVec<Mat*>& ports_value, const TVec<Mat*>& ports_gradient) Adapt based on the output gradient: this method should only be called just after a corresponding fprop; it should be called with the same arguments as fprop for the first two arguments (and output should not have been modified since then).

Since sub-classes are supposed to learn ONLINE, the object is 'ready-to-be-used' just after any bpropUpdate. N.B. The DEFAULT IMPLEMENTATION JUST CALLS bpropUpdate(input, output, input_gradient, output_gradient) AND IGNORES INPUT GRADIENT.

Reimplemented from PLearn::OnlineLearningModule.

Definition at line 82 of file SquaredErrModule.cc.

References PLearn::OnlineLearningModule::input_size, PLearn::OnlineLearningModule::output_size, PLWARNING, PLearn::TVec< T >::size(), and target_size.

Referenced by bbpropUpdate().

{
    int in_size = input.size();
    int out_size = output.size();
    int og_size = output_gradient.size();

    // size check
    if( in_size != input_size + target_size )
    {
        PLWARNING("SquaredErrModule::bpropUpdate: 'input.size()' should be\n"
                  " equal to 'input_size' + 'target_size' (%i != %i + %i)\n",
                  in_size, input_size, target_size);
    }
    if( out_size != output_size )
    {
        PLWARNING("SquaredErrModule::bpropUpdate: output.size()' should be\n"
                  " equal to 'output_size' (%i != %i)\n",
                  out_size, output_size);
    }
    if( og_size != output_size )
    {
        PLWARNING("SquaredErrModule::bpropUpdate: 'output_gradient.size()'\n"
                  " should be equal to 'output_size' (%i != %i)\n",
                  og_size, output_size);
    }
}

Here is the call graph for this function:

Here is the caller graph for this function:

void PLearn::SquaredErrModule::bpropUpdate ( const Vec input,
const Vec output,
Vec input_gradient,
const Vec output_gradient 
) [virtual]

Definition at line 112 of file SquaredErrModule.cc.

References i, PLearn::OnlineLearningModule::input_size, PLearn::OnlineLearningModule::output_size, PLERROR, PLearn::TVec< T >::resize(), PLearn::TVec< T >::size(), PLearn::TVec< T >::subVec(), and target_size.

{
    int in_size = input.size();
    int out_size = output.size();
    int og_size = output_gradient.size();
    bool is_final_cost = false; // if yes, output_gradient is 1

    // size check
    if( in_size != input_size + target_size )
    {
        PLERROR("SquaredErrModule::bpropUpdate: 'input.size()' should be\n"
                " equal to 'input_size' + 'target_size' (%i != %i + %i)\n",
                in_size, input_size, target_size);
    }
    if( out_size != output_size )
    {
        PLERROR("SquaredErrModule::bpropUpdate: output.size()' should be\n"
                " equal to 'output_size' (%i != %i)\n",
                out_size, output_size);
    }
    if( og_size == 0 )
    {
        /*
        PLWARNING("SquaredErrModule::bpropUpdate: you are not providing"
                  "output_gradient.\n"
                  "Assuming this is the final cost, and output_gradient=1.\n");
         */
        is_final_cost = true;
    }
    else if( og_size != output_size )
    {
        PLERROR("SquaredErrModule::bpropUpdate: 'output_gradient.size()'\n"
                " should be equal to 'output_size' (%i != %i)\n",
                og_size, output_size);
    }

    Vec input_ = input.subVec( 0, input_size );
    Vec target = input.subVec( input_size, target_size );
    input_gradient.resize( input_size );
    for( int i=0 ; i<input_size ; i++ )
    {
        if( is_final_cost )
            input_gradient[i] = 2*( input_[i] - target[i] );
        else
            input_gradient[i] = 2*( input_[i] - target[i] )*output_gradient[0];
    }
}

Here is the call graph for this function:

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

Post-constructor.

The normal implementation should call simply inherited::build(), then this class's build_(). This method should be callable again at later times, after modifying some option fields to change the "architecture" of the object.

Reimplemented from PLearn::OnlineLearningModule.

Definition at line 253 of file SquaredErrModule.cc.

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

Here is the call graph for this function:

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

This does the actual building.

Reimplemented from PLearn::OnlineLearningModule.

Definition at line 271 of file SquaredErrModule.cc.

References PLearn::OnlineLearningModule::input_size, PLearn::OnlineLearningModule::output_size, PLWARNING, and target_size.

Referenced by build().

{
    if( input_size < 0 )
    {
        PLWARNING("SquaredErrModule::build_: 'input_size' is < 0.\n"
                  "You should set it to a positive integer.\n"
                  "Defaulting to '1' (scalar version).");
        input_size = 1;
    }
    if( output_size != 1 )
    {
        PLWARNING("SquaredErrModule::build_: 'output_size' (%i) should be 1.\n"
                  "Setting 'output_size' to 1.\n", output_size);
        output_size = 1;
    }

    target_size = input_size;
}

Here is the caller graph for this function:

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

Reimplemented from PLearn::Object.

Definition at line 53 of file SquaredErrModule.cc.

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

Declares the class options.

Reimplemented from PLearn::OnlineLearningModule.

Definition at line 266 of file SquaredErrModule.cc.

References PLearn::OnlineLearningModule::declareOptions().

Here is the call graph for this function:

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

Reimplemented from PLearn::OnlineLearningModule.

Definition at line 102 of file SquaredErrModule.h.

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

Reimplemented from PLearn::OnlineLearningModule.

Definition at line 53 of file SquaredErrModule.cc.

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

reset the parameters to the state they would be BEFORE starting training.

Note that this method is necessarily called from build().

Implements PLearn::OnlineLearningModule.

Definition at line 246 of file SquaredErrModule.cc.

{
//    target = Vec( input_size );
}
void PLearn::SquaredErrModule::fprop ( const Vec input,
Vec output 
) const [virtual]

given the input, compute the output (possibly resize it appropriately) SOON TO BE DEPRECATED, USE fprop(const TVec<Mat*>& ports_value)

Reimplemented from PLearn::OnlineLearningModule.

Definition at line 64 of file SquaredErrModule.cc.

References PLearn::OnlineLearningModule::input_size, PLearn::OnlineLearningModule::output_size, PLERROR, PLearn::TVec< T >::resize(), PLearn::TVec< T >::size(), PLearn::TVec< T >::subVec(), PLearn::sumsquare(), and target_size.

{
    int in_size = input.size();
    // size check
    if( in_size != input_size+target_size )
    {
        PLERROR("SquaredErrModule::fprop: 'input.size()' should be equal\n"
                " to 'input_size' + 'target_size' (%i != %i + %i)\n",
                in_size, input_size, target_size);
    }

    Vec target = input.subVec( input_size, target_size );
    Vec input_ = input.subVec( 0, input_size );
    output.resize( output_size );
    output[0] = sumsquare( input_ - target );
}

Here is the call graph for this function:

OptionList & PLearn::SquaredErrModule::getOptionList ( ) const [virtual]

Reimplemented from PLearn::Object.

Definition at line 53 of file SquaredErrModule.cc.

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

Reimplemented from PLearn::Object.

Definition at line 53 of file SquaredErrModule.cc.

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

Reimplemented from PLearn::Object.

Definition at line 53 of file SquaredErrModule.cc.

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

Transforms a shallow copy into a deep copy.

Reimplemented from PLearn::OnlineLearningModule.

Definition at line 259 of file SquaredErrModule.cc.

References PLearn::OnlineLearningModule::makeDeepCopyFromShallowCopy().

{
    inherited::makeDeepCopyFromShallowCopy(copies);

//    deepCopyField(target, copies);
}

Here is the call graph for this function:


Member Data Documentation

Reimplemented from PLearn::OnlineLearningModule.

Definition at line 102 of file SquaredErrModule.h.

Definition at line 114 of file SquaredErrModule.h.

Referenced by bbpropUpdate(), bpropUpdate(), build_(), and fprop().


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