SPIDER The Spider Objects

DATA object

   Data object
   Stores data into two components X (input) and Y (output).
   An object is created with:
     data(X,Y); (no name)
     data(X); (in case there's no output, e.g clustering)
   Public attributes:
    X          -- input matrix        
    Y          -- output matrix       
    index      -- original parent indices of examples 
    findex     -- original parent indices of features 
   Public methods:
    d=get(data,index,featInd)             -- return data,index only (ind)examples,(find)features 
    x=get_x(data,index,fInd)              -- return x,index only (ind)examples,(find)features 
    y=get_y(data,index,fInd)              -- return y,index only (ind)examples,(find)features 
    d=set_x(data,X,indes,featInd)         -- set inputs indexed by (ind)examples,(find)features
    d=set_y(data,Y,index,featInd)         -- set outputs indxed by (ind)examples,(find)features
    [indes,featInd]=get_index(d)          -- returns example and feature indices
    [numEx,vDim,oDim,numCls]=get_dim(d)   -- returns number of examples,features,output dimensions,classes
   Import methods:
    The Data object can import libsvm or arff files for file exchange with weka/libsvm.  
    Note that arff does not encode what is input or output thus the resulting data object has an empty Y member.
    Nominal attributes are mapped to an index value and missing attributes are indicated by "NaN".
    get_x(data([1:5;6:10;11:15]),[1 3],[3:5])
     There is more than one data available.  See also
        help data/data.m