addpath('./helper/');
workdir = './matfiles/';
heat_savedir = [workdir 'heat_kernels_q/'];
acc_savedir  = [workdir 'acc/'];


%% --- Prep For CV -------------------------------------------------------------
%maxNumCachedPerClass = 4000;
%numFolds             = 20;
%numTrainPerClass     = 250;
%numTestPerClass      = 100;
%c1labels = {'M11\d*','M12\d*','M13\d*','M14\d*'};
%load([workdir 'topics.mat']);
%load([workdir 'member.7k.mat']);
%
%% class1 will have label '1' while class2 will have label '-1'
%[ids trn_samps tst_samps trn_zs tst_zs] = prepOneVsRest( c1labels, topics, member, ...
%		maxNumCachedPerClass, numFolds, numTrainPerClass, numTestPerClass );
%
%clear topics member;
%% save all variables here
%save([workdir 'OneVsRest.7k.mat']); 
load([workdir 'OneVsRest.7k.mat']); 



%% --- Initialization ----------------------------------------------------------
load([workdir 'tf.7k.mat']);
load([workdir 'doclen.7k.mat']);	% a row vector equiv to sum(tc,1)
tf = tf(:,ids);			% prune unused documents (cols)
doclen = doclen(ids);	% prune unused documents (cols)

%% remove unused words (rows)
%% this is not necessary since it would only affect the transD2 computation,
%%   which never divides by the doc-freq of words
%inactivewords = find(~any(tf,2));
%tf(inactivewords,:) = [];   % (rows)
%tc(inactivewords,:) = [];   % (rows)


filelist = dir(heat_savedir);
ptrn = 'Hc(\d+)t(\d+)';

trbfgamma = [.125 .25 .5 1 2];

% --- Loop through paramters to test translation kernel -----------------------
fprintf('CV RBF Kernel under translation\n');
ii = 0;
for fln = {filelist.name}
	file = char(fln);
	match = regexp(file,ptrn,'tokens');
	% jvd: by re-checking file existence, we can control execution by removing,
	%      but not adding kernels
	if numel(match)==0 | ~exist([heat_savedir file],'file')
		continue;
	end

	ii = ii+1;
	fprintf('%s\n',file);

	c = str2num(match{1}{1})/100;
	t = str2num(match{1}{2})/100;
	load([heat_savedir file]);
	params(:,ii) = [c;t];	% consolidating

	% --- Optional Feature selection (on T) -----------------------------------
	% H(inactivewords,inactivewords) = [];	% not needed, see above
	% jvd: make sure normalize last!
	T = normalizeRows(H);

	% cache the expected dists^2 using only relevant docs (pruned above) and
	% words the occur
	[ED2 A] = transD2tf(tf,doclen,T);

	% --- RBF Kernel ----------------------------------------------------------
	for jj=1:length(trbfgamma)
		fprintf('\ttrbfgamma = %f\n',trbfgamma(jj));
		% no need for specfic flags, except maybe: "-c 15.0" or something
		TAcc = svmCVOneVsRest( trn_samps, tst_samps, trn_zs, tst_zs, ...
					tf, numFolds, '', 'tranrbf', ED2, A, trbfgamma(jj) );
		tacc(:,jj,ii) = TAcc;
	end
end


% --- off the shelf RBF kernel ------------------------------------------------
fprintf('CV RBF Kernel\n');
rbfgamma = [.05 .08 .125 .25 .375 .5 .625 .75 .875 1 2];
for ii=1:length(rbfgamma)
	fprintf('\trbfgamma = %f\n',rbfgamma(ii));
	% jvd: TODO: fix svmCVOneVsRest so that it takes a gamma for stndrbf
	TAcc = svmCVOneVsRest( trn_samps, tst_samps, trn_zs, tst_zs, ...
				tf, numFolds, '', 'stndrbf', [], [], rbfgamma(ii) );
	acc(:,ii)  = Acc;
end

clear_except tacc acc trbfgamma rbfgamma params workdir;
save([workdir 'acc_svm_rbf_q.7k.mat']);