%CENSUS Weighted census income data
%PRTools UCI dataset import, 199523+99762 objects, 41 mixed features, 2 classes
%
% [TRAIN,TEST] = CENSUS(VAL)
% TRAIN_TEST = CENSUS(VAL)
%
%DESCRIPTION
%This command downloads one of the UCI data sets, converts it into PRTools
%format and stores it locally for future use. Consult the related website
%for further information. Please make the appropriate references in
%publications that make use of this dataset.
%
%This dataset contains a number of categorical features with N > 2
%categories. They may be converted to N new real features by CAT2REAL.
%
%Dataset has missing values. By default all objects with missing values are
%removed. Use VAL=NaN to avoid this. For other options see MISVAL.
%
%SEE ALSO PRTools Guide, UCI Website
%PRTOOLS, DATASETS, SETFEATDOM, FEATTYPES, CAT2REAL, MISVAL,ADULT
% Copyright: R.P.W. Duin
function varargout = census(val)
if nargin<1
val = 'remove';
end
varargout = cell(1,nargout);
[varargout{:}] = pr_loadmatfile;
if isempty(varargout{1})
opt.format = 'ncnncnccccccccccnnncccccncccccncccccncnnnc';
opt1.size = 6;
opt2.size = 3;
opt.dsetname = 'Census Income KDD';
% no matfiles found, create them
[varargout{:}] = pr_download_uci('Census-Income+(KDD)',{'census-income.data.gz','census-income.test.gz'},{opt1,opt2,opt});
end
varargout = varargout*misval(val);