%SETDAT Reset data and feature labels of dataset for classification output % % A = SETDAT(A,DATA,W) % % INPUT % A Dataset % DATA Dataset or double % W Mapping (optional) % % OUTPUT % A Dataset % % DESCRIPTION % The data in the dataset A is replaced by DATA (dataset or double). The % number of objects in A and DATA should be equal. Optionally, A is given % the feature labels and the output size as defined by the the mapping W. % This call is identical to: % A = SETDATA(A,DATA); % A = SET(A,'featlab',GETLABELS(W),'featsize',GETSIZE_OUT(W)); % % SEE ALSO (PRTools Guide) % PRDATASET, SETDATA % Copyright: R.P.W. Duin, duin@ph.tn.tudelft.nl % Faculty of Applied Sciences, Delft University of Technology % P.O. Box 5046, 2600 GA Delft, The Netherlands % $Id: setdat.m,v 1.2 2006/03/08 22:06:58 duin Exp $ function a = setdat(a,b,w) if ~isdataset(a) a = prdataset(a); end a = setdata(a,b); if (nargin > 2) if (~isa(w,'prmapping')) error('Third parameter should be mapping') end % Special case: 2 classes, p(class2) = 1-p(class1). %No! this should not be done here! %if (size(a,2) == 1) & (size(w,2) == 2) % a = [a 1-a]; %end % Add attributes of W to A. a = setfeatsize(a,getsize_out(w)); featlab = getlabels(w); if ~isempty(featlab); a = setfeatlab(a,featlab); end %a = set(a,'featlab',getlabels(w),'featsize',getsize_out(w)); end return;