a = load('lung.dat');
b = max(a,[],2) - min(a,[],2);
indexf = find(b > 1.0);
m = median(a,2);
indexm = find(m > 6.5);
ind = intersect(indexf, indexm);
ind = intersect(ind,63:length(a)); %remove 62 controls
data = a(ind,:);
data = data'; %samples in rows and gene in columns;
save data.txt -ascii -tabs data;

import dataframe.*;
des = TextArray('h133+.genedescriptors','\t','na',0);
filtered = des.selectRowsm(ind);
filtered.save('filtered.des',false,false,true);
txt2mat('filtered.des');

ClinInfo = TextArray('Lung_fixed.clinical.txt','\t', 'NaN', 1);

%celltype response
CellType = ClinInfo.columnm('CellType');
A = find(CellType == 'A');
type = zeros(1,91);
type(A) = 1;
Type = TextArray(type,[],[],'#');
Type.save('type.txt',false,false,true);

%weight: use all samples
weight = ones(1,91);
Weight = TextArray(weight,[],[],'#');
Weight.save('weight.txt',false,false,true);

%survival time
clindata = ClinInfo.data;
survivaltime = clindata(:,7);
SurvTime = TextArray(survivaltime,[],[]);
SurvTime.save('survtime.txt',false,false,true);

%censorship
censor = clindata(:,8);
%translate: 1->0 for censored, 3->1 for observed
c1 = find(censor == 1);
c3 = find(censor == 3);
censor(c1) = 0;
censor(c3) = 1;
Censor = TextArray(censor,[],[],'#');
Censor.save('censor.txt',false,false,true);