Thin C++ wrapper class around C implementation of LWPR. More...
#include <lwpr.hh>
Public Member Functions | |
| LWPR_Object (int nIn, int nOut) | |
| Creates an LWPR_Object for the given input and output dimensionalities. | |
| LWPR_Object (const LWPR_Object &otherObj) | |
| Creates a copy of an LWPR_Object. | |
| LWPR_Object (const char *filename) | |
| Creates an LWPR_Object from a binary file, or if compiled with support for EXPAT, an XML file. | |
| ~LWPR_Object () | |
| Destroys an LWPR_Object and disposes allocated memory. | |
| int | writeXML (const char *filename) |
| Write the model to an XML file. | |
| int | writeBinary (const char *filename) |
| Write the model to a binary file. | |
| doubleVec | update (const doubleVec &x, const doubleVec &y) |
| Updates an LWPR model with a given input/output pair (x,y). | |
| doubleVec | predict (const doubleVec &x, double cutoff=0.001) |
| Computes the prediction of an LWPR model given an input vector x. | |
| doubleVec | predict (const doubleVec &x, doubleVec &confidence, double cutoff=0.001) |
| Computes the prediction of an LWPR model given an input vector x. Also computes confidence bounds per output dimension. | |
| doubleVec | predict (const doubleVec &x, doubleVec &confidence, doubleVec &maxW, double cutoff=0.001) |
| Computes the prediction of an LWPR model given an input vector x. Also computes confidence bounds and maximal activation per output dimension. | |
| void | setInitD (double delta) |
| Sets a spherical initial distance metric. | |
| void | setInitD (const doubleVec &initD) |
| Sets a diagonal or full initial distance metric. | |
| void | setInitAlpha (double alpha) |
| Sets init_alpha (learning rate for 2nd order distance metric updates). | |
| void | wGen (double w_gen) |
| Sets w_gen (threshold for adding new receptive fields). | |
| void | wPrune (double w_prune) |
| Sets w_prune (threshold for removing a receptive field). | |
| void | penalty (double pen) |
| Sets penalty (pre-factor for smoothing term in distance metric updates). | |
| void | initLambda (double iLam) |
| Sets initial forgetting factor. | |
| void | tauLambda (double tLam) |
| Sets annealing rate for forgetting factor. | |
| void | finalLambda (double fLam) |
| Sets final forgetting factor. | |
| void | initS2 (double init_s2) |
| Sets initial value for covariance computation SSs2. | |
| void | updateD (bool update) |
| Determines whether distance matrix updates are to be performed. | |
| void | diagOnly (bool dOnly) |
| Determines whether distance matrices should be treaded as diagonal-only. | |
| void | useMeta (bool meta) |
| Determines whether 2nd order distance matrix updates are to be performed. | |
| void | metaRate (double rate) |
| Sets the learning rate for 2nd order distance matrix updates. | |
| void | kernel (LWPR_Kernel kern) |
| Sets the kernel to be used in the LWPR model. | |
| void | kernel (const char *str) |
| Sets the kernel (either "Gaussian" or "BiSquare") to be used in the LWPR model. | |
| int | nData () const |
| Returns the number of training data the model has seen. | |
| int | nIn () const |
| Returns the input dimensionality. | |
| int | nOut () const |
| Returns the output dimensionality. | |
| double | wGen () const |
| Returns w_gen (threshold for adding new receptive fields). | |
| double | wPrune () const |
| Returns w_prune (threshold for removing a receptive field). | |
| double | penalty () const |
| Returns penalty (pre-factor for smoothing term in distance metric updates). | |
| double | initLambda () const |
| Returns initial forgetting factor. | |
| double | tauLambda () const |
| Returns annealing rate for forgetting factor. | |
| double | finalLambda () const |
| Returns final forgetting factor. | |
| double | initS2 () const |
| Returns initial value for the covariance computation SSs2. | |
| bool | updateD () |
| Returns whether distance matrix updates are performed. | |
| bool | diagOnly () |
| Returns whether distance matrices are treaded as diagonal-only. | |
| bool | useMeta () |
| Returns whether 2nd order distance matrix updates are performed. | |
| double | metaRate () |
| Returns learning rate for 2nd order distance matrix updates. | |
| LWPR_Kernel | kernel () |
| Returns the kernel. | |
| doubleVec | meanX () |
| Returns the mean of all input samples the model has seen. | |
| doubleVec | varX () |
| Returns the variance of all input samples the model has seen. | |
| void | normIn (const doubleVec &norm) |
| Sets the input normalisation (expected scale or standard deviation of input data. | |
| doubleVec | normIn () const |
| Returns the input normalisation factors. | |
| void | normOut (const doubleVec &norm) |
| Sets the output normalisation (expected scale or standard deviation of output data. | |
| doubleVec | normOut () const |
| Returns the output normalisation factors. | |
| int | numRFS (int outDim) |
| Returns the number of receptive fields for output dimension "outDim". | |
| std::vector< int > | numRFS () |
| Returns the number of receptive fields for all output dimensions. | |
| LWPR_ReceptiveFieldObject | getRF (int outDim, int index) const |
| Returns a wrapper object for inspecting a receptive field. | |
Data Fields | |
| LWPR_Model | model |
| Underlying C structure. | |
Thin C++ wrapper class around C implementation of LWPR.
| LWPR_Object::LWPR_Object | ( | int | nIn, | |
| int | nOut | |||
| ) | [inline] |
Creates an LWPR_Object for the given input and output dimensionalities.
| nIn | Desired input dimensionality | |
| nOut | Desired output dimensionality |
In case there is insufficient memory for allocating the underlying LWPR_Model (C library), an OUT_OF_MEMORY exception is thrown.
References lwpr_init_model(), model, and LWPR_Exception::OUT_OF_MEMORY.
| LWPR_Object::LWPR_Object | ( | const LWPR_Object & | otherObj | ) | [inline] |
Creates a copy of an LWPR_Object.
| otherObj | LWPR_Object to be duplicated. |
In case there is insufficient memory for allocating the underlying LWPR_Model (C library), an OUT_OF_MEMORY exception is thrown.
References lwpr_duplicate_model(), model, and LWPR_Exception::OUT_OF_MEMORY.
| LWPR_Object::LWPR_Object | ( | const char * | filename | ) | [inline] |
Creates an LWPR_Object from a binary file, or if compiled with support for EXPAT, an XML file.
| filename | Name of file to read the model from |
In case there are problems with reading the file, an IO_ERROR exception is thrown.
References LWPR_Exception::IO_ERROR, lwpr_read_binary(), lwpr_read_xml(), and model.
| LWPR_ReceptiveFieldObject LWPR_Object::getRF | ( | int | outDim, | |
| int | index | |||
| ) | const [inline] |
Returns a wrapper object for inspecting a receptive field.
| outDim | Desired output dimension | |
| index | Index of the receptive field within that output dimension |
| LWPR_Exception::OUT_OF_RANGE | if the parameters outDim and index are out of range |
Please note that the retrieved object is only guaranteed to be valid as long as the underlying LWPR model is not changed, i.e. predictions are fine, but updates may result in pruning just the receptive field this wrapper points to!!!
References model, LWPR_Model::nOut, LWPR_SubModel::numRFS, LWPR_Exception::OUT_OF_RANGE, LWPR_SubModel::rf, and LWPR_Model::sub.
| doubleVec LWPR_Object::predict | ( | const doubleVec & | x, | |
| doubleVec & | confidence, | |||
| doubleVec & | maxW, | |||
| double | cutoff = 0.001 | |||
| ) | [inline] |
Computes the prediction of an LWPR model given an input vector x. Also computes confidence bounds and maximal activation per output dimension.
| [in] | x | Input vector |
| [out] | confidence | Vector to store the confidence bounds, will be resized if necessary |
| [out] | maxW | Vector to store maximum activations, will be resized if necessary |
| [in] | cutoff | A threshold parameter (default = 0.001). Receptive fields with activation below the cutoff are ignored |
| LWPR_Exception::BAD_INPUT_DIM | if the parameter x does not match the model dimensions |
References LWPR_Exception::BAD_INPUT_DIM, lwpr_predict(), model, LWPR_Model::nIn, and LWPR_Model::nOut.
| doubleVec LWPR_Object::predict | ( | const doubleVec & | x, | |
| doubleVec & | confidence, | |||
| double | cutoff = 0.001 | |||
| ) | [inline] |
Computes the prediction of an LWPR model given an input vector x. Also computes confidence bounds per output dimension.
| [in] | x | Input vector |
| [out] | confidence | Vector to store the confidence bounds, will be resized if necessary |
| [in] | cutoff | A threshold parameter (default = 0.001). Receptive fields with activation below the cutoff are ignored |
| LWPR_Exception::BAD_INPUT_DIM | if the parameter x does not match the model dimensions |
References LWPR_Exception::BAD_INPUT_DIM, lwpr_predict(), model, LWPR_Model::nIn, and LWPR_Model::nOut.
Computes the prediction of an LWPR model given an input vector x.
| x | Input vector. | |
| cutoff | A threshold parameter (default = 0.001). Receptive fields with activation below the cutoff are ignored |
| LWPR_Exception::BAD_INPUT_DIM | if the parameter x does not match the model dimensions |
References LWPR_Exception::BAD_INPUT_DIM, lwpr_predict(), model, LWPR_Model::nIn, and LWPR_Model::nOut.
| void LWPR_Object::setInitD | ( | const doubleVec & | initD | ) | [inline] |
Sets a diagonal or full initial distance metric.
| initD | Either a vector with nIn elements, containing the diagonal of the desired distance metric, or a vector with nIn*nIn elements containing the complete distance matrix. |
| LWPR_Exception::BAD_INIT_D | if the parameter initD gives rise to a non-positive matrix |
References LWPR_Exception::BAD_INIT_D, LWPR_Exception::BAD_INPUT_DIM, lwpr_set_init_D(), lwpr_set_init_D_diagonal(), model, and LWPR_Model::nIn.
| void LWPR_Object::setInitD | ( | double | delta | ) | [inline] |
Sets a spherical initial distance metric.
| delta | Width parameter, distance matrix will be delta * eye(nIn) |
| LWPR_Exception::BAD_INIT_D | if the parameter delta is <= 0, giving rise to a non-positive matrix |
References LWPR_Exception::BAD_INIT_D, lwpr_set_init_D_spherical(), and model.
Updates an LWPR model with a given input/output pair (x,y).
| x | Input vector | |
| y | Output vector |
| LWPR_Exception::OUT_OF_MEMORY | if a receptive field would have to be added, but memory could not be allocated | |
| LWPR_Exception::BAD_INPUT_DIM | if the parameter x does not match the model dimensions | |
| LWPR_Exception::BAD_OUTPUT_DIM | if the parameter y does not match the model dimensions |
References LWPR_Exception::BAD_INPUT_DIM, LWPR_Exception::BAD_OUTPUT_DIM, lwpr_update(), model, LWPR_Model::nIn, LWPR_Model::nOut, and LWPR_Exception::OUT_OF_MEMORY.
| int LWPR_Object::writeBinary | ( | const char * | filename | ) | [inline] |
Write the model to a binary file.
| filename | Name of the file, which will we overwritten if it already exists |
References lwpr_write_binary(), and model.
| int LWPR_Object::writeXML | ( | const char * | filename | ) | [inline] |
Write the model to an XML file.
| filename | Name of the file, which will we overwritten if it already exists |
References lwpr_write_xml(), and model.
1.6.3