Revamp ann modules

src/mlpack/methods/ann/convolution_rules/naive_convolution.hpp

@@ -104,7 +104,15 @@ class NaiveConvolution
         filter.n_cols - 1 + input.n_cols - 1) = input;
 
     NaiveConvolution<ValidConvolution>::Convolution(inputPadded, filter,
+<<<<<<< HEAD
+<<<<<<< HEAD
         output, 1, 1);
+=======
+        output, dW, dH);
+>>>>>>> Use the stride parameter inside the convolution function.
+=======
+        output, 1, 1);
+>>>>>>> Remove stride paramater from svd and fft convolution rule.
   }
 
   /*

src/mlpack/methods/ann/layer/add.hpp

@@ -39,7 +39,18 @@ class Add
    *
    * @param outSize The number of output units.
    */
+<<<<<<< HEAD
+<<<<<<< HEAD
   Add(const size_t outSize);
+=======
+  Add(const size_t outSize) : outSize(outSize)
+  {
+    weights.set_size(outSize, 1);
+  }
+>>>>>>> Refactor ann layer.
+=======
+  Add(const size_t outSize);
+>>>>>>> Split layer modules into definition and implementation.
 
   /**
    * Ordinary feed forward pass of a neural network, evaluating the function
@@ -49,7 +60,18 @@ class Add
    * @param output Resulting output activation.
    */
   template<typename eT>
+<<<<<<< HEAD
+<<<<<<< HEAD
+  void Forward(const arma::Mat<eT>&& input, arma::Mat<eT>&& output);
+=======
+  void Forward(const arma::Mat<eT>&& input, arma::Mat<eT>&& output)
+  {
+    output = input + weights;
+  }
+>>>>>>> Refactor ann layer.
+=======
   void Forward(const arma::Mat<eT>&& input, arma::Mat<eT>&& output);
+>>>>>>> Split layer modules into definition and implementation.
 
   /**
    * Ordinary feed backward pass of a neural network, calculating the function
@@ -63,7 +85,18 @@ class Add
   template<typename eT>
   void Backward(const arma::Mat<eT>&& /* input */,
                 const arma::Mat<eT>&& gy,
+<<<<<<< HEAD
+<<<<<<< HEAD
                 arma::Mat<eT>&& g);
+=======
+                arma::Mat<eT>&& g)
+  {
+    g = gy;
+  }
+>>>>>>> Refactor ann layer.
+=======
+                arma::Mat<eT>&& g);
+>>>>>>> Split layer modules into definition and implementation.
 
   /*
    * Calculate the gradient using the output delta and the input activation.
@@ -75,7 +108,18 @@ class Add
   template<typename eT>
   void Gradient(const arma::Mat<eT>&& /* input */,
                 arma::Mat<eT>&& error,
+<<<<<<< HEAD
+<<<<<<< HEAD
+                arma::Mat<eT>&& gradient);
+=======
+                arma::Mat<eT>&& gradient)
+  {
+    gradient = error;
+  }
+>>>>>>> Refactor ann layer.
+=======
                 arma::Mat<eT>&& gradient);
+>>>>>>> Split layer modules into definition and implementation.
 
   //! Get the parameters.
   OutputDataType const& Parameters() const { return weights; }
@@ -106,7 +150,18 @@ class Add
    * Serialize the layer
    */
   template<typename Archive>
+<<<<<<< HEAD
+<<<<<<< HEAD
   void Serialize(Archive& ar, const unsigned int /* version */);
+=======
+  void Serialize(Archive& ar, const unsigned int /* version */)
+  {
+    ar & data::CreateNVP(weights, "weights");
+  }
+>>>>>>> Refactor ann layer.
+=======
+  void Serialize(Archive& ar, const unsigned int /* version */);
+>>>>>>> Split layer modules into definition and implementation.
 
  private:
   //! Locally-stored number of output units.
@@ -131,7 +186,16 @@ class Add
 } // namespace ann
 } // namespace mlpack
 
+<<<<<<< HEAD
+<<<<<<< HEAD
+// Include implementation.
+#include "add_impl.hpp"
+
+=======
+>>>>>>> Refactor ann layer.
+=======
 // Include implementation.
 #include "add_impl.hpp"
 
+>>>>>>> Split layer modules into definition and implementation.
 #endif

src/mlpack/methods/ann/layer/add_merge.hpp

@@ -38,7 +38,18 @@ class AddMerge
 {
  public:
   //! Create the AddMerge object.
+<<<<<<< HEAD
+<<<<<<< HEAD
   AddMerge();
+=======
+  AddMerge()
+  {
+    // Nothing to do here.
+  }
+>>>>>>> Refactor ann layer.
+=======
+  AddMerge();
+>>>>>>> Split layer modules into definition and implementation.
 
   /**
    * Ordinary feed forward pass of a neural network, evaluating the function
@@ -48,7 +59,23 @@ class AddMerge
    * @param output Resulting output activation.
    */
   template<typename InputType, typename OutputType>
+<<<<<<< HEAD
+<<<<<<< HEAD
+  void Forward(const InputType&& /* input */, OutputType&& output);
+=======
+  void Forward(const InputType&& /* input */, OutputType&& output)
+  {
+    output = boost::apply_visitor(outputParameterVisitor, network.front());
+
+    for (size_t i = 1; i < network.size(); ++i)
+    {
+      output += boost::apply_visitor(outputParameterVisitor, network[i]);
+    }
+  }
+>>>>>>> Refactor ann layer.
+=======
   void Forward(const InputType&& /* input */, OutputType&& output);
+>>>>>>> Split layer modules into definition and implementation.
 
   /**
    * Ordinary feed backward pass of a neural network, calculating the function
@@ -62,7 +89,18 @@ class AddMerge
   template<typename eT>
   void Backward(const arma::Mat<eT>&& /* input */,
                 arma::Mat<eT>&& gy,
+<<<<<<< HEAD
+<<<<<<< HEAD
                 arma::Mat<eT>&& g);
+=======
+                arma::Mat<eT>&& g)
+  {
+    g = gy;
+  }
+>>>>>>> Refactor ann layer.
+=======
+                arma::Mat<eT>&& g);
+>>>>>>> Split layer modules into definition and implementation.
 
   /*
    * Add a new module to the model.
@@ -106,7 +144,18 @@ class AddMerge
    * Serialize the layer.
    */
   template<typename Archive>
+<<<<<<< HEAD
+<<<<<<< HEAD
+  void Serialize(Archive& ar, const unsigned int /* version */);
+=======
+  void Serialize(Archive& ar, const unsigned int /* version */)
+  {
+    ar & data::CreateNVP(network, "network");
+  }
+>>>>>>> Refactor ann layer.
+=======
   void Serialize(Archive& ar, const unsigned int /* version */);
+>>>>>>> Split layer modules into definition and implementation.
 
  private:
   std::vector<LayerTypes> network;
@@ -133,7 +182,16 @@ class AddMerge
 } // namespace ann
 } // namespace mlpack
 
+<<<<<<< HEAD
+<<<<<<< HEAD
+// Include implementation.
+#include "add_merge_impl.hpp"
+
+=======
+>>>>>>> Refactor ann layer.
+=======
 // Include implementation.
 #include "add_merge_impl.hpp"
 
+>>>>>>> Split layer modules into definition and implementation.
 #endif