Betweeness centrality has been removed : it was making things much slower and...

Betweeness centrality has been removed : it was making things much slower and wasn't needed right now. We will enable it when we need it
main.ccp has been split into 3 files :
    graph.cpp for graph generation and manipulation
    results.cpp to gather the results of the simulation
    main.cpp to launch the simulation and display the results

simulation/graph.cpp

+#include "main.h"
+
+Graph::Graph(int size, int k, int maxPeers, mt19937 rng) : 
+    distrib(uniform_int_distribution<int>(0, size-1)),
+    size(size)
+{
+    adjacency = new vector<int>[size];
+    for(int i=0; i<size; i++)
+    {
+        set<int> alreadyConnected;
+        alreadyConnected.insert(i);
+
+        for(int j=0; j<k; j++)
+        {
+            int otherNode;
+
+            while(alreadyConnected.count(otherNode = distrib(rng)) == 1 
+                || otherNode > i && adjacency[otherNode].size() > maxPeers-10
+                || adjacency[otherNode].size() > maxPeers) 
+            { }
+            adjacency[i].push_back(otherNode);
+            adjacency[otherNode].push_back(i);
+        }
+    }
+}
+
+void Graph::GetDistancesFrom(int node, int* distance)
+{
+    for(int j=0; j<size; j++)
+        distance[j] = -1;
+    distance[node] = 0;
+
+    queue<int> remainingNodes;
+    remainingNodes.push(node);
+
+    while(!remainingNodes.empty())
+    {
+        int node = remainingNodes.front();
+        remainingNodes.pop();
+
+        for(int neighbor : adjacency[node])
+            if(distance[neighbor] == -1)
+            {
+                distance[neighbor] = distance[node]+1;
+                remainingNodes.push(neighbor);
+            }
+    }
+}
+
+// kill the last proportion*size machines of the graph
+void Graph::KillMachines(float proportion)
+{
+    // TODO
+}

simulation/main.cpp

 // To compile with -std=c++0x
-// The GET_BC option might not be working
-//#define GET_BC // Uncomment this line to get the betweeness centrality
 #include "main.h"
 
 int n = 1000; // the number of peer
 int k = 10; // each peer try to make k connections with the others
 int maxPeer = 25; // no more that 25 connections per peer
-int runs = 100;  // use this to run the simulation multiple times to get more accurate values
-int betweenessDiv = 20; // to kown how to sample the BC. Max BC should be betweenessDiv*100
-
-Graph::Graph(int size, int k, int maxPeers, mt19937 rng) : 
-    distrib(uniform_int_distribution<int>(0, n-1)),
-    size(size)
-{
-    adjacency = new vector<int>[size];
-    for(int i=0; i<size; i++)
-    {
-        set<int> alreadyConnected;
-        alreadyConnected.insert(i);
-
-        for(int j=0; j<k; j++)
-        {
-            int otherNode;
-
-            while(alreadyConnected.count(otherNode = distrib(rng)) == 1 
-                || otherNode > i && adjacency[otherNode].size() > maxPeers-10
-                || adjacency[otherNode].size() > maxPeers) 
-            { }
-            adjacency[i].push_back(otherNode);
-            adjacency[otherNode].push_back(i);
-        }
-    }
-}
-
-void Graph::GetDistancesFrom(int node, int* distance)
-{
-    for(int j=0; j<size; j++)
-        distance[j] = -1;
-    distance[node] = 0;
-
-    queue<int> remainingNodes;
-    remainingNodes.push(node);
-
-    while(!remainingNodes.empty())
-    {
-        int node = remainingNodes.front();
-        remainingNodes.pop();
-
-        for(int neighbor : adjacency[node])
-            if(distance[neighbor] == -1)
-            {
-                distance[neighbor] = distance[node]+1;
-                remainingNodes.push(neighbor);
-            }
-    }
-}
-
-// kill the last proportion*size machines of the graph
-void Graph::KillMachines(float proportion)
-{
-    // TODO
-}
+int runs = 10;  // use this to run the simulation multiple times to get more accurate values
 
 int main()
 {
     mt19937 rng(time(NULL));
-
-    // Init the parameters
-    double array(arityDistrib, maxPeer+1);
-    double array(distanceDistrib, n);
-    double array(betweenessDistrib, 100);
-    int disconnected = 0;
+    Results results(maxPeer, 20);
 
     for(int r=0; r<runs; r++)
     {
@@ -78,104 +17,35 @@ int main()
         cout.flush();
 
         Graph graph(n, k, maxPeer, rng);
-        double array(betweeness, n);
-
-        // Get the arity distribution
-        for(int i=0; i<n; i++)
-            arityDistrib[graph.adjacency[i].size()]++;
+        results.UpdateArity(graph);
 
         // Compute the shortest path
-        // TODO : optimise this
-        // switch to int64 ?
         for(int i=0; i<graph.size; i++)
         {
             int distance[graph.size];
-            // if(i%10==0) cout << "Computing distances from node " << i << endl;
             graph.GetDistancesFrom(i, distance);
-
-            // retrieve the distance
-            int maxDistance = -1;
-            for(int j=0; j<graph.size; j++)
-                if(distance[j] != -1)
-                {
-                    maxDistance = max(distance[j], maxDistance);
-                    distanceDistrib[distance[j]]++;
+            results.UpdateDistance(distance, graph.size);
         }
-                else
-                    disconnected++;
-
-#ifdef GET_BC
-            // Get the betweeness
-            double toBePushed[graph.size];
-            for(int j=0; j<n; j++)
-                toBePushed[j] = 1;
-          
-            // TODO : change this  into a true sort ?
-            // run accross the nodes in the right order
-            // we don't need to sort them since we will only run across them a few times
-            for(int d=maxDistance; d>=0; d--)
-                for(int j=0; j<graph.size; j++)
-                    if(distance[j] == d)
-                    {
-                        int nMin, min = -1;
-                        for(int neighbor : graph.adjacency[j])
-                        {
-                            if(distance[neighbor] < min || min == -1)
-                            {
-                                min = distance[neighbor];
-                                nMin = 1;
-                            }
-                            else if(distance[neighbor] == min)
-                                nMin++;
-                        }
-                        
-                        double singleScore = toBePushed[j]/nMin;
-                        for(int neighbor : graph.adjacency[j])
-                            if(distance[neighbor] == min)
-                                toBePushed[neighbor] += singleScore;
-                      
-                        betweeness[j] += toBePushed[j] - 1;
-                    }
-#endif
-        }
-#ifdef GET_BC
-        // Get the betweeness distribution
-        for(int i=0; i<n; i++)
-            betweenessDistrib[min((int)betweeness[i]/betweenessDiv, 99)]++;
-#endif
     }
     cout << "\r            \r";
 
+    results.Finalise();
+
     // Display the parameters we have mesured
     cout << "Arity :" << endl;
-    for(int i=0; i<=maxPeer; i++)
-        if(arityDistrib[i] != 0)
-        {
-            arityDistrib[i] /= (double)(n*runs);
-            cout << i << " : " << arityDistrib[i] << endl;
-        }
+    for(int i=0; i<=results.maxArity; i++)
+        if(results.arityDistrib[i] != 0)
+            cout << i << " : " << results.arityDistrib[i] << endl;
     
     cout << "Distance :" << endl;
     double nLinks = n*(n-1)*runs;
-    for(int i=0; i<n; i++)
-        if(distanceDistrib[i] != 0)
-        {
-            distanceDistrib[i] /= nLinks - disconnected;
-            cout << i << " : " << distanceDistrib[i] << endl;
-        }
+    for(int i=0; i<= results.maxDistance; i++)
+        if(results.distanceDistrib[i] != 0)
+            cout << i << " : " << results.distanceDistrib[i] << endl;
     
     cout << "Probability that a node is not reachable : " 
-         << disconnected/nLinks 
-         << " (" << disconnected << " total)" << endl;
-
-#ifdef GET_BC
-    cout << "Betweeness :" << endl;
-    double nNodes = n*runs;
-    for(int i=0; i<100; i++)
-        if(betweenessDistrib[i] != 0)
-            cout << betweenessDiv*i << " -> " << betweenessDiv*(i+1) << " : " 
-                 << betweenessDistrib[i]/nNodes << endl;
-#endif
+         << results.disconnectionProba
+         << " (" << results.disconnected << " total)" << endl;
 
     cout << endl;
     return 0;

simulation/main.h

@@ -4,11 +4,6 @@
 #include <queue>
 #include <set>
 
-#define max(a,b) a>b?a:b
-#define min(a,b) a<b?a:b
-#define clamp(a,b,c) max(a,min(b, c))
-#define array(name, size) name[size]; for(int i=0; i<size; i++) name[i]=0;
-
 using namespace std;
 
 class Graph
@@ -26,3 +21,28 @@ private:
     uniform_int_distribution<int> distrib;
 };
 
+class Results
+{
+public:
+    Results(int maxArity, int maxDistance);
+    ~Results();
+
+    void UpdateArity(const Graph& graph);
+    void UpdateDistance(int* distance, int nSamples);
+    void Finalise();
+
+    double* arityDistrib;
+    double* distanceDistrib;
+    double disconnectionProba;
+    double arityTooBig;
+    double distanceTooBig;
+    int64_t disconnected;
+    int64_t nAritySample;
+    int64_t nDistanceSample;
+    int maxArity;
+    int maxDistance;
+
+private:
+    void AddAritySample(int arity);
+    void AddDistanceSample(int distance);
+};

simulation/results.cpp

+#include "main.h"
+
+Results::Results(int maxArity, int maxDistance) : 
+    maxArity(maxArity), maxDistance(maxDistance)
+{
+    arityDistrib = new double[maxArity+1];
+    for(int i=0; i<=maxArity; i++)
+        arityDistrib[i] = 0;
+
+    distanceDistrib = new double[maxDistance+1];
+    for(int i=0; i<=maxDistance; i++)
+        distanceDistrib[i] = 0;
+
+    nAritySample = 0;
+    nDistanceSample = 0;
+    arityTooBig = 0;
+    distanceTooBig = 0;
+    disconnected = 0;
+}
+
+Results::~Results()
+{
+    delete[] arityDistrib;
+    delete[] distanceDistrib;
+}
+
+void Results::UpdateArity(const Graph& graph)
+{
+    for(int i=0; i<graph.size; i++)
+        AddAritySample(graph.adjacency[i].size());
+}
+
+void Results::UpdateDistance(int* distance, int nSamples)
+{
+    for(int i=0; i<nSamples; i++)
+        AddDistanceSample(distance[i]);
+}
+
+void Results::AddAritySample(int arity)
+{
+    if(arity <= maxArity)
+        arityDistrib[arity]++;
+    else
+        distanceTooBig++;
+    nAritySample++;
+}
+
+void Results::AddDistanceSample(int distance)
+{
+    if(distance == -1)
+        disconnected++;
+    else if(distance <= maxDistance)
+        distanceDistrib[distance]++;
+    else
+        distanceTooBig++;
+    nDistanceSample++;
+}
+
+void Results::Finalise()
+{
+    for(int i=0; i<=maxArity; i++)
+        arityDistrib[i] /= nAritySample;
+    for(int i=0; i<=maxDistance; i++)
+        distanceDistrib[i] /= nDistanceSample;
+    disconnectionProba = ((double)disconnected)/nDistanceSample;
+    distanceTooBig/= nDistanceSample;
+    arityTooBig /= nAritySample;
+}