perClass Documentation
version 5.0 (21-sep-2016)
 SDDATA Constructor for data set object

      A=SDDATA(DATA)
      A=SDDATA(DATA,LAB)

      A=SDDATA(CELL)   % create sddata from a cell array
      A=SDDATA(CELL1D,LIST) % create nominal feature using values in LIST
      A=SDDATA(LAB,LIST)    % create nominal feature using values in LIST

 INPUT
    DATA     Data matrix samples times features (double precision)
    LAB      Sample labels (SDLAB object, char array or cell array)

 OUTPUT
    A        SDDATA object

 DESCRIPTION
 SDDATA object represents a data set. Data matrix DATA contains samples
 in rows and features in columns. When sample labels LAB are not
 provided, SDDATA will assign all samples to 'unknown' class.

 SDDATA may be created from a CELL array. The string columns will be
 converted to nominal features. DATA.feattype stores the nominam/real
 feature type. The unique values of a nominal feature are stored in
 DATA.featlist feature property, values in DATA are the indices to the
 featlist. To convert new nominal values to this numerical representation
 use featlist specific to the nominal feature SDDATA(CELL1D,LIST)

 BASIC OPERATIONS
  +A                       return data matrix (shorthand for double(A))
  A(1:10)                  return with first 10 samples
  A(:,[2 4])               return all samples and features 2 and 4
  A(:,:,3)                 return 3rd class
  A(:,:,{'apple','lemon'}) return subset with named classes
  A.lab                    return sample labels
  A.featlab                return feature labels
  A.patient=sdlab(patname) create patient labels

 Get featlist for a nominal feature:
  LIST=A.featlist{ind}
 Convert a 1D cell-array of strings into nominal feature:
  A=SDDATA(CELL1D,LIST)

 SEE ALSO
 SDLAB, GETLAB, SETLAB, SUBSET, RANDSUBSET