Complete and fix KDB

sample/main.cc

@@ -221,41 +221,44 @@ int main(int argc, char** argv)
     cout << endl;
     cout << "Class name: " << className << endl;
     // Build Network
-    auto network = bayesnet::Network(1.0);
+    // auto network = bayesnet::Network(1.0);
-    build_network(network, network_name, maxes);
+    // build_network(network, network_name, maxes);
-    network.fit(Xd, y, features, className);
+    // network.fit(Xd, y, features, className);
-    cout << "Hello, Bayesian Networks!" << endl;
+    // cout << "Hello, Bayesian Networks!" << endl;
-    showNodesInfo(network, className);
+    // showNodesInfo(network, className);
-    //showCPDS(network);
+    // //showCPDS(network);
-    cout << "Score: " << network.score(Xd, y) << endl;
+    // cout << "Score: " << network.score(Xd, y) << endl;
-    cout << "PyTorch version: " << TORCH_VERSION << endl;
+    // cout << "PyTorch version: " << TORCH_VERSION << endl;
-    cout << "BayesNet version: " << network.version() << endl;
+    // cout << "BayesNet version: " << network.version() << endl;
-    unsigned int nthreads = std::thread::hardware_concurrency();
+    // unsigned int nthreads = std::thread::hardware_concurrency();
-    cout << "Computer has " << nthreads << " cores." << endl;
+    // cout << "Computer has " << nthreads << " cores." << endl;
-    cout << "****************** First ******************" << endl;
+    // cout << "****************** First ******************" << endl;
-    auto metrics = bayesnet::Metrics(network.getSamples(), features, className, network.getClassNumStates());
+    // auto metrics = bayesnet::Metrics(network.getSamples(), features, className, network.getClassNumStates());
-    cout << "conditionalEdgeWeight " << endl;
+    // cout << "conditionalEdgeWeight " << endl;
-    auto conditional = metrics.conditionalEdgeWeights();
+    // auto conditional = metrics.conditionalEdgeWeights();
-    cout << conditional << endl;
+    // cout << conditional << endl;
-    long m = features.size() + 1;
+    // long m = features.size() + 1;
-    auto matrix = torch::from_blob(conditional.data(), { m, m });
+    // auto matrix = torch::from_blob(conditional.data(), { m, m });
-    cout << matrix << endl;
+    // cout << matrix << endl;
-    cout << "****************** Second ******************" << endl;
+    // cout << "****************** Second ******************" << endl;
-    auto metrics2 = bayesnet::Metrics(Xd, y, features, className, network.getClassNumStates());
+    // auto metrics2 = bayesnet::Metrics(Xd, y, features, className, network.getClassNumStates());
-    cout << "conditionalEdgeWeight " << endl;
+    // cout << "conditionalEdgeWeight " << endl;
-    auto conditional2 = metrics2.conditionalEdgeWeights();
+    // auto conditional2 = metrics2.conditionalEdgeWeights();
-    cout << conditional2 << endl;
+    // cout << conditional2 << endl;
-    long m2 = features.size() + 1;
+    // long m2 = features.size() + 1;
-    auto matrix2 = torch::from_blob(conditional2.data(), { m, m });
+    // auto matrix2 = torch::from_blob(conditional2.data(), { m, m });
-    cout << matrix2 << endl;
+    // cout << matrix2 << endl;
     cout << "****************** KDB ******************" << endl;
     map<string, vector<int>> states;
     for (auto feature : features) {
         states[feature] = vector<int>(maxes[feature]);
     }
     states[className] = vector<int>(maxes[className]);
-    auto kdb = bayesnet::KDB(1);
+    auto kdb = bayesnet::KDB(2);
     kdb.fit(Xd, y, features, className, states);
+    for (auto line : kdb.show()) {
+        cout << line << endl;
+    }
     cout << "****************** KDB ******************" << endl;
     return 0;
 }

src/BaseClassifier.cc

@@ -8,7 +8,7 @@ namespace bayesnet {
     BaseClassifier& BaseClassifier::build(vector<string>& features, string className, map<string, vector<int>>& states)
     {
 
-        dataset = torch::cat({ X, y.view({150, 1}) }, 1);
+        dataset = torch::cat({ X, y.view({y.size(0), 1}) }, 1);
         this->features = features;
         this->className = className;
         this->states = states;
@@ -86,4 +86,8 @@ namespace bayesnet {
         Tensor y_pred = predict(X);
         return (y_pred == y).sum().item<float>() / y.size(0);
     }
+    void BaseClassifier::show()
+    {
+        model.show();
+    }
 }

src/BaseClassifier.h

@@ -28,8 +28,8 @@ namespace bayesnet {
         BaseClassifier& fit(vector<vector<int>>& X, vector<int>& y, vector<string>& features, string className, map<string, vector<int>>& states);
         Tensor predict(Tensor& X);
         float score(Tensor& X, Tensor& y);
+        void show();
     };
-
 }
 #endif
 

src/KDB.cc

@@ -12,7 +12,7 @@ namespace bayesnet {
         sort(indices.begin(), indices.end(), [&nums](int i, int j) {return nums[i] > nums[j];});
         return indices;
     }
-    KDB::KDB(int k, float theta = 0.03) : BaseClassifier(Network()), k(k), theta(theta) {}
+    KDB::KDB(int k, float theta) : BaseClassifier(Network()), k(k), theta(theta) {}
     void KDB::train()
     {
         /*
@@ -76,23 +76,39 @@ namespace bayesnet {
     {
         auto n_edges = min(k, static_cast<int>(S.size()));
         auto cond_w = clone(weights);
+        cout << "Conditional edge weights cloned for idx " << idx << endl;
+        cout << cond_w << endl;
         bool exit_cond = k == 0;
         int num = 0;
         while (!exit_cond) {
-            auto max_minfo = argmax(cond_w.index({ "...", idx })).item<int>();
+            auto max_minfo = argmax(cond_w.index({ idx, "..." })).item<int>();
             auto belongs = find(S.begin(), S.end(), max_minfo) != S.end();
             if (belongs && cond_w.index({ idx, max_minfo }).item<float>() > theta) {
                 try {
-                    model.addEdge(features[idx], features[max_minfo]);
+                    model.addEdge(features[max_minfo], features[idx]);
                     num++;
                 }
                 catch (const invalid_argument& e) {
                     // Loops are not allowed
                 }
             }
-            cond_w.index_put_({ "...", max_minfo }, -1);
+            cond_w.index_put_({ idx, max_minfo }, -1);
-            auto candidates = cond_w.gt(theta);
+            cout << "Conditional edge weights cloned for idx " << idx << " After -1" << endl;
+            cout << cond_w << endl;
+            cout << "cond_w.index({ idx, '...'})" << endl;
+            cout << cond_w.index({ idx, "..." }) << endl;
+            auto candidates_mask = cond_w.index({ idx, "..." }).gt(theta);
+            auto candidates = candidates_mask.nonzero();
+            cout << "Candidates mask" << endl;
+            cout << candidates_mask << endl;
+            cout << "Candidates: " << endl;
+            cout << candidates << endl;
+            cout << "Candidates size: " << candidates.size(0) << endl;
             exit_cond = num == n_edges || candidates.size(0) == 0;
         }
     }
+    vector<string> KDB::show()
+    {
+        return model.show();
+    }
 }

src/KDB.h

@@ -10,9 +10,10 @@ namespace bayesnet {
         float theta;
         void add_m_edges(int idx, vector<int>& S, Tensor& weights);
     protected:
-        void train();
+        void train() override;
     public:
-        KDB(int k, float theta);
+        KDB(int k, float theta = 0.03);
+        vector<string> show();
     };
 }
 #endif

src/Network.cc

@@ -245,5 +245,18 @@ namespace bayesnet {
         }
         return result;
     }
+    vector<string> Network::show()
+    {
+        vector<string> result;
+        // Draw the network
+        for (auto node : nodes) {
+            string line = node.first + " -> ";
+            for (auto child : node.second->getChildren()) {
+                line += child->getName() + ", ";
+            }
+            result.push_back(line);
+        }
+        return result;
+    }
 
 }

src/Network.h

@@ -44,6 +44,7 @@ namespace bayesnet {
         torch::Tensor conditionalEdgeWeight();
         vector<vector<double>> predict_proba(const vector<vector<int>>&);
         double score(const vector<vector<int>>&, const vector<int>&);
+        vector<string> show();
         inline string version() { return "0.1.0"; }
     };
 }