docTransformV
Centre for Vision, Speech & Signal Processing
 

Transforming V Data using a NumFuncC

It is frequently necessary when given a NumVDataSetC to process all the input vectors using a class derived from NumFuncC, eg after feature selection or reduction.

This can be achieved very easily using the following segment of code which illustrates the case with labelled feature vectors: 

     NumVLDataSetC samples;       // data set that is provided
     NumFuncC transform;          // some object derived from NumFuncC that
                                  // has already be setup or trained
     samples = NumVLDataSetC(transform.Evaluate(samples.InputSet()),samples.OutputSet);
The last line uses the Evaluate member function to process the list of input vectors in samples. A new data set is then constructed using the transformed input vectors list and the original output values list from the sample set.

Author: Robert Crida, Created: 12/6/1998, Generated by DocCreate: March 20, 2001