wolffd@0: function [relative, absolute, test_cn, train_cn] =  sim_get_traintest_clip_overlap(datafile)
wolffd@0: %
wolffd@0: %  get_traintest_clip_overlap(datafile)
wolffd@0: %
wolffd@0: % returns the percentage of overlapping constraints 
wolffd@0: % with the corresponding training set for each test set
wolffd@0: % 
wolffd@0: % how many percent of the test set are reappearing in the training set
wolffd@0: 
wolffd@0: 
wolffd@0: % simdata = load(datafile);
wolffd@0: if nargin < 1
wolffd@0:     simdata = load('comp_partBinData_unclustered_cupaper_01');
wolffd@0: else
wolffd@0:     simdata = load(datafile);
wolffd@0: end
wolffd@0: nTestSets = size(simdata.partBinTst, 2); % num cv bins
wolffd@0: ntrainsizes = size(simdata.partBinTrn, 2); % num increases of training
wolffd@0: 
wolffd@0: 
wolffd@0: absolute = zeros(nTestSets, ntrainsizes);
wolffd@0: relative = zeros(nTestSets, ntrainsizes);
wolffd@0: for k = 1:nTestSets % all test/training combinatios
wolffd@0:     
wolffd@0:     % get clips of this test set
wolffd@0:     test_clips = unique([simdata.partBinTst{k}(:,1); simdata.partBinTst{k}(:,2); simdata.partBinTst{k}(:,3)]);
wolffd@0:     test_cn(k) = numel(test_clips);
wolffd@0:     for m = 1:ntrainsizes
wolffd@0:         
wolffd@0:         % get clips of this training set
wolffd@0:         train_clips = unique([simdata.partBinTrn{k,m}(:,1); simdata.partBinTrn{k,m}(:,2); simdata.partBinTrn{k,m}(:,3)]);
wolffd@0:         
wolffd@0:         % intersect both clip sets
wolffd@0:         same = intersect(train_clips, test_clips);
wolffd@0:         
wolffd@0:         % get stats
wolffd@0:         absolute(k,m) = numel(same);
wolffd@0:         relative(k,m) = absolute(k,m) / numel(test_clips);
wolffd@0:     end
wolffd@0:     train_cn(k) = numel(train_clips);
wolffd@0: end