SAMI/MissingNodes_SrcCode2014/CheckProperties2.m

clear;
clc;

%affinity calculation types
% affinity_calculation_shortest_path = 0;
% affinity_calculation_euclid = 1;
% affinity_calculation_common_friends = 2;
% affinity_calculation_random_clustering = 3;
% affinity_calculation_adamic_adar = 4;
% affinity_calculation_katz_beta_0_5 = 5;
% affinity_calculation_katz_beta_0_05 = 6;
% affinity_calculation_katz_beta_0_005 = 7;

affinities = 3; %%[2,3,4,6]; %3,4]; %,6];
num_missing_nodes_arr = 11; %%[11 21 31 41 50]; %%5:5:30; %10:10:50; %%[11 21 31 41 50]; %10:10:50;
percentKnownPHsVec = 1;

datasetDir = 'D:/SocialNets/Steam/Results_Test/FF75_Steam_GM2_20/' ;
results_dir = strcat(datasetDir,'All_W3/');
filePrefix = 'Steam_*';
netSizes = 2048; %%[2048];
runAlgFlag = 1;
debugFlag = 0;
dumpFlag = 0;
numThreshold = 0;
%%maxAttStat = 0.25; % use this attribute only if it appears less than this percentage
numAttrCols = 21; %40; %50
%attSelected = ones(1,numAttrCols);
%attSelected(1) = 0; %skip country
attWeight = 0.3;
addMissingAtt = 0;

date_now = clock;
date_now = strcat(num2str(date_now(1)),'_',num2str(date_now(2)),'_', num2str(date_now(3)),'_', num2str(date_now(4)), num2str(date_now(5)),'_', num2str(date_now(6)));
LogMsg(sprintf('%s Strating CheckProperties ...',date_now));

outFile = sprintf('%sPropertiesData_%s.txt', results_dir, date_now);
fileID = 0;

for nodes = netSizes
    
    prefix = sprintf('%s%s%d_%s',datasetDir,filePrefix,nodes,'*.mat');
    files = dir(prefix);
    firstIter = 1;
    
    for iter = 1 %1:2 %100         % loop over same network with different missing nodes
        
        for i = 1:size(files,1)     % loop over the list of networks
            
            file = files(i).name;
            
            fprintf('reading network information from file %s%s ...\n', datasetDir, file);
            load(strcat(datasetDir, file), 'data');
            data = sparse(data);
            
            for maxAttStat = [0.25 0.3 0.4 0.5 0.75 1]
                
                attSelected = ones(1,numAttrCols);
                %skip all games
                %                 for g=2:1:11
                %                     attSelected(g) = 0;
                %                 end
                %skip all groups
                %                 for g=12:1:50
                %                     attSelected(g) = 0;
                %                 end
                attSelectedNum = nnz(attSelected);
                
                attFile = strrep(file, '.txt.mat', '.usr.txt');
                [attributes, attUpperRange, attSelected, attStat] = PrepareAttributes2(datasetDir, attFile, nodes, numAttrCols, maxAttStat, attSelected);
                
                if runAlgFlag == 1
                    date_now = clock;
                    date_now = strcat(num2str(date_now(1)),'_',num2str(date_now(2)),'_', num2str(date_now(3)),'_', num2str(date_now(4)), num2str(date_now(5)),'_', num2str(date_now(6)));
                    
                    % make sure dump & results directories exist
                    if (firstIter == 1 && i == 1)
                        firstIter = 0;
                        if isdir(results_dir) == 0
                            mkdir(results_dir);
                        end
                        fileID = fopen(outFile,'w');
                        fprintf(fileID,'\tfile\ti\titer\tnodes\tedges\tnonedges\tmaxAttStat\tattSelectedNum\tnumAtt');
                        fprintf(fileID,'\tedgesAttNNZ\tedgesAttOne\tpercAttOne\tedgesAttZero\tpercAttZero\tedgesAvg\tedgesAvgNNZ');
                        fprintf(fileID,'\tnonedgesAttNNZ\tnonedgesAttOne\tpercAttOne\tnonedgesAttZero\tpercAttZero\tnonedgesAvg\tnonedgesAvgNNZ');
                        fprintf(fileID,'\n');
                    end
                    
                    % pre process attributes (translate to binary attributes)
                    [attData, totalAttNum] = PreProcessDataAttributes(data, attributes, attUpperRange);
                    % Calc Attributes Affinity - similarity score
                    fprintf('calculating attribtes affinity matrix...\n');
                    [attAffinity] = CalcAttributesAffinity(data, attData, nodes, addMissingAtt, 1, 1);
                    
                    numAtt = nnz(attSelected); %%size(attributes,2);
                    edges = nnz(data);
                    nonedges = nodes*nodes-nodes-edges;
                    
                    % save iteration data
                    fprintf(fileID,'\t%s',file);
                    fprintf(fileID,'\t%d',i);
                    fprintf(fileID,'\t%d',iter);
                    fprintf(fileID,'\t%d',nodes);
                    fprintf(fileID,'\t%d',edges);
                    fprintf(fileID,'\t%d',nonedges);
                    fprintf(fileID,'\t%d',maxAttStat);
                    fprintf(fileID,'\t%d',attSelectedNum);
                    fprintf(fileID,'\t%d',numAtt);
                    
                    % edges properties
                    edgesInd = find(data);
                    edgesAttInd = find(attAffinity(edgesInd));
                    edgesAttNNZ = size(edgesAttInd,1);
                    edgesAttOne = size(find(attAffinity(edgesInd)==1),1);
                    edgesAttZero = edges-edgesAttNNZ;
                    edgesAttSum = full(sum(sum(attAffinity(edgesInd))));
                    edgesAvg = edgesAttSum/edges;
                    edgesAvgNNZ = edgesAttSum/edgesAttNNZ;
                    
                    fprintf(fileID,'\t%d',edgesAttNNZ);
                    fprintf(fileID,'\t%d\t%d',edgesAttOne,edgesAttOne/edges);
                    fprintf(fileID,'\t%d\t%d',edgesAttZero,edgesAttZero/edges);
                    fprintf(fileID,'\t%d',edgesAvg);
                    fprintf(fileID,'\t%d',edgesAvgNNZ);
                    
                    % nonedges properties
                    nonedges = nodes*nodes-nodes-edges;
                    nonedgesInd = find(data==0);
                    nonedgesAttInd = find(attAffinity(nonedgesInd));
                    nonedgesAttNNZ = size(nonedgesAttInd,1)-nodes;
                    nonedgesAttOne = size(find(attAffinity(nonedgesInd)==1),1)-nodes;
                    nonedgesAttZero = nonedges-nonedgesAttNNZ;
                    nonedgesAttSum = full(sum(sum(attAffinity(nonedgesInd))))-nodes;
                    nonedgesAvg = nonedgesAttSum/nonedges;
                    nonedgesAvgNNZ = nonedgesAttSum/nonedgesAttNNZ;
                    
                    fprintf(fileID,'\t%d',nonedgesAttNNZ);
                    fprintf(fileID,'\t%d\t%d',nonedgesAttOne,nonedgesAttOne/nonedges);
                    fprintf(fileID,'\t%d\t%d',nonedgesAttZero,nonedgesAttZero/nonedges);
                    fprintf(fileID,'\t%d',nonedgesAvg);
                    fprintf(fileID,'\t%d',nonedgesAvgNNZ);
                    fprintf(fileID,'\n');
                    
                    %LogMsg(sprintf('Results for file %s,iter %d at %s',file,iter,out_file));
                    %fprintf('Completed RunExperiment cycle - results at %s.\n',out_file);
                end
            end
            
            %  beep;
        end
    end
end

if fileID ~= 0
    fclose(fileID);
end

LogMsg(sprintf('%s Completed RunExperiment.',date_now));