Revert "Update ConMod.m file"

This reverts commit d6188f9.

Author: WPZgithub

ConMod.m

@@ -3,21 +3,23 @@
 % modules in multiple networks
 %
 % INPUT:
-%   multiNetworks : a cell contains multiple networks, each of which is
-%                   presented by edgelist format or a full matrix
-%                   with N nodes
+%   multiNetworks: a cell contains multiple networks, each is  
+%                  presented by a sparse matrix or a full matrix with N nodes
 %   N: the number of all nodes
 %   K: the number of hidden factors
-%   lambda: a vector which contains the parameters for balancing the relative
+%   lambda: a vector containing the parameters for balancing the relative
 %		   weight among different views
 %   xita:  the parameter for selecting nodes
 %   maxIter:  the maximum number of iterations for multi-view NMF
 %
 % OUTPUT:
-%   modulesfinal: a cell which contains the final conserved modules
+%   modulesfinal: a cell contains the final conserved modules
 %
 % Peizhuo Wang ([email protected])
 
+%% parameters
+
+
 %% Calculting the feature matrices
 disp('Calculating the strengh matrix and the uniformity matrix...')
 [Strength, Distribution] = featureNets(multiNetworks, N);

README.md

@@ -5,7 +5,7 @@
 %   X: the adjacency matrix of a network
 %   Hc: initialization for consensus factor matrix 
 %   H: initialization for factor matrix of each view
-%   lambda: a vector which contains the parameters for balancing the relative
+%   lambda: a vector containing the parameters for balancing the relative
 %           weight among different views
 %   MaxIter: the maximal number of iterations for alternating minimization
 %   epsilon: the convergence parameter

SNMFforView.m

@@ -10,6 +10,7 @@
 %       FPR: the False Positive Rate
 %       Accuracy:
 %       MCC: the Matthews Correlation Coefficient
+%       I: Confusion Matrix
 %
 % Peizhuo Wang ([email protected])
 

evaluation.m

@@ -2,8 +2,8 @@
 % Compute two feature metrices from multiple networks
 %
 % INPUT:
-%   multiNetworks : a cell contains multiple networks, each of which is
-%                   presented by edgelist format or a full matrix
+%   multiNetworks : a cell contains multiple networks, each is
+%                   presented by a sparse matrix or a full matrix
 %                   with N nodes
 %   N : the number of all nodes
 %
@@ -19,7 +19,7 @@
 Strength = zeros(N);
 temp = zeros(N);
 A = zeros(N);
-if (m <= 3) % Edgelist format
+if (m <= 3) % Sparse matrix format
     for k = 1:network_count
         theMatrix = multiNetworks{k};
         [edge_count, col_count] = size(theMatrix);
@@ -73,9 +73,9 @@
 Participation = (network_count/(network_count-1)) * (1-(temp./(A.^2)));
 Participation(isinf(Participation)) = 0;
 Participation(isnan(Participation)) = 0;
-Participation = Participation - diag(diag(Participation)); % The diagonal is 0
+Participation = Participation - diag(diag(Participation));
 
 Strength = A./network_count;
-Strength = Strength - diag(diag(Strength)); % The diagonal is 0
+Strength = Strength - diag(diag(Strength));
 
 end

featureNets.m

@@ -44,7 +44,7 @@
 % 
 % % Selecting nodes from the consensus factors
 % xita = 1.5;
-% modules_final = moduleNodesSelection( Hc, xita );
+% modules_final = modulesTruing( Hc, xita );
 % runtime = toc;
 % disp(['Running time: ', num2str(runtime), ' sec.'])
 

main_run.m

@@ -1,13 +1,13 @@
 function [ modulesFinal ] = moduleNodesSelection( Hc, xita )
-%  Assigning the module members by a soft node selection procedure
-% and then truing the modules to obtain more accurate results
+%  A soft node selection procedure from the consensus factors to assign the module members 
+% and then truing the  modules to obtain more accurate results
 %
 % INPUT:
 %   Hc: the consensus factor matrix
 %   xita: the parameter for selecting nodes
 %
 % OUTPUT:
-%   modulesFinal: a cell which contains the final result modules
+%   modulesFinal: a cell contains the final result modules
 %
 % Peizhuo Wang ([email protected])
 
@@ -26,7 +26,7 @@
 
 for i = 1:size(HPI, 1)-1
     for j = (i+1):size(HPI, 2)
-        if HPI(i,j)>0.5 % merge these two modules
+        if HPI(i,j)>0.5
             [Y, I] = max([moduleSignal(i), moduleSignal(j)]);
             if I == 1
                 modulesFinal{j} = [];
@@ -43,7 +43,6 @@
     end
 end
 
-% Only modules with no less than 5 nodes are kept
 i = 1;
 while i ~= length(modulesFinal)+1
     if isempty(modulesFinal{i}) || (length(modulesFinal{i})<5)

moduleNodesSelection.m

@@ -2,14 +2,14 @@
 % Multi-View Non-negative symmetric Matrix Factorization
 %
 % INPUT:
-%   X: a cell which contains symmetric matrices
+%   X: a cell containing symmetric matrices
 %   K: the number of hidden factors
-%   lambda: a vector which contains the parameters for balancing the relative
+%   lambda: a vector containing the parameters for balancing the relative
 %                weight among different views
 %   maxiter: the maximum number of iterations
 %
 % OUTPUT:
-%   H: a cell which contains factor matrices for all views
+%   H: a cell containing factor matrices for all views
 %   Hc: the result consensus factor matrix
 %   objValue: the value of objective function
 %
@@ -86,9 +86,9 @@
         obj_consensus = norm(H{i} - Hc, 'fro')^2;
         obj = obj + obj_body + lambda(i)*obj_consensus;
 
-%         errX = mean(mean(abs(obj_body)))/mean(mean(X{i}));
-%         errH = mean(mean(abs(obj_consensus)))/mean(mean(H{i}));
-%         err = errX + errH;
+        errX = mean(mean(abs(obj_body)))/mean(mean(X{i}));
+        errH = mean(mean(abs(obj_consensus)))/mean(mean(H{i}));
+        err = errX + errH;
     end