Apply @rcurtin comments

Signed-off-by: Omar Shrit <omar@avontech.fr>

src/mlpack/methods/dbscan/dbscan.hpp

@@ -47,13 +47,14 @@ namespace mlpack {
  *      with.
  */
 template<typename RangeSearchType = RangeSearch<>,
-         typename PointSelectionPolicy = OrderedPointSelection,
-         typename MatType = arma::mat>
+         typename PointSelectionPolicy = OrderedPointSelection>
 class DBSCAN
 {
  public:
+   //! Easy access to the MatType.
+  typedef typename RangeSearchType::Mat MatType;
   //! Easy access to Element Type of the matrix.
-  typedef typename MatType::elem_type ElemType;
+  typedef typename RangeSearchType::Mat::elem_type ElemType;
   /**
    * Construct the DBSCAN object with the given parameters.  The batchMode
    * parameter should be set to false in the case where RAM issues will be
@@ -114,9 +115,6 @@ class DBSCAN
   //! Maximum distance between two points to be part of same cluster.
   ElemType epsilon;
 
-  //! Zero, just a variable holder for zero value, can be f16, f32 or double.
-  ElemType zero = 0.0;
-
   //! Minimum number of points to be in the epsilon-neighborhood (including
   //! itself) for the point to be a core-point.
   size_t minPoints;

src/mlpack/methods/dbscan/dbscan_impl.hpp

@@ -19,9 +19,8 @@ namespace mlpack {
 /**
  * Construct the DBSCAN object with the given parameters.
  */
-template<typename RangeSearchType, typename PointSelectionPolicy,
-         typename MatType>
-DBSCAN<RangeSearchType, PointSelectionPolicy, MatType>::DBSCAN(
+template<typename RangeSearchType, typename PointSelectionPolicy>
+DBSCAN<RangeSearchType, PointSelectionPolicy>::DBSCAN(
     const ElemType epsilon,
     const size_t minPoints,
     const bool batchMode,
@@ -40,9 +39,8 @@ DBSCAN<RangeSearchType, PointSelectionPolicy, MatType>::DBSCAN(
  * Performs DBSCAN clustering on the data, returning number of clusters
  * and also the centroid of each cluster.
  */
-template<typename RangeSearchType, typename PointSelectionPolicy,
-         typename MatType>
-size_t DBSCAN<RangeSearchType, PointSelectionPolicy, MatType>::Cluster(
+template<typename RangeSearchType, typename PointSelectionPolicy>
+size_t DBSCAN<RangeSearchType, PointSelectionPolicy>::Cluster(
     const MatType& data,
     MatType& centroids)
 {
@@ -58,9 +56,8 @@ size_t DBSCAN<RangeSearchType, PointSelectionPolicy, MatType>::Cluster(
  * Performs DBSCAN clustering on the data, returning number of clusters,
  * the centroid of each cluster and also the list of cluster assignments.
  */
-template<typename RangeSearchType, typename PointSelectionPolicy,
-         typename MatType>
-size_t DBSCAN<RangeSearchType, PointSelectionPolicy, MatType>::Cluster(
+template<typename RangeSearchType, typename PointSelectionPolicy>
+size_t DBSCAN<RangeSearchType, PointSelectionPolicy>::Cluster(
     const MatType& data,
     arma::Row<size_t>& assignments,
     MatType& centroids)
@@ -94,9 +91,8 @@ size_t DBSCAN<RangeSearchType, PointSelectionPolicy, MatType>::Cluster(
  * Performs DBSCAN clustering on the data, returning the number of clusters and
  * also the list of cluster assignments.
  */
-template<typename RangeSearchType, typename PointSelectionPolicy,
-         typename MatType>
-size_t DBSCAN<RangeSearchType, PointSelectionPolicy, MatType>::Cluster(
+template<typename RangeSearchType, typename PointSelectionPolicy>
+size_t DBSCAN<RangeSearchType, PointSelectionPolicy>::Cluster(
     const MatType& data,
     arma::Row<size_t>& assignments)
 {
@@ -146,9 +142,8 @@ size_t DBSCAN<RangeSearchType, PointSelectionPolicy, MatType>::Cluster(
  * and can save on RAM usage.  It may be slower than the batch search with a
  * dual-tree algorithm.
  */
-template<typename RangeSearchType, typename PointSelectionPolicy,
-         typename MatType>
-void DBSCAN<RangeSearchType, PointSelectionPolicy, MatType>::PointwiseCluster(
+template<typename RangeSearchType, typename PointSelectionPolicy>
+void DBSCAN<RangeSearchType, PointSelectionPolicy>::PointwiseCluster(
     const MatType& data,
     UnionFind& uf)
 {
@@ -182,7 +177,7 @@ void DBSCAN<RangeSearchType, PointSelectionPolicy, MatType>::PointwiseCluster(
     visited[index] = true;
 
     // Do the range search for only this point.
-    rangeSearch.Search(data.col(index), RangeType<ElemType>(zero, epsilon),
+    rangeSearch.Search(data.col(index), RangeType<ElemType>(ElemType(0.0), epsilon),
         neighbors,
         distances);
 
@@ -226,9 +221,8 @@ void DBSCAN<RangeSearchType, PointSelectionPolicy, MatType>::PointwiseCluster(
  * and also the list of cluster assignments.  This can perform search in batch,
  * naive search).
  */
-template<typename RangeSearchType, typename PointSelectionPolicy,
-         typename MatType>
-void DBSCAN<RangeSearchType, PointSelectionPolicy, MatType>::BatchCluster(
+template<typename RangeSearchType, typename PointSelectionPolicy>
+void DBSCAN<RangeSearchType, PointSelectionPolicy>::BatchCluster(
     const MatType& data,
     UnionFind& uf)
 {
@@ -237,7 +231,7 @@ void DBSCAN<RangeSearchType, PointSelectionPolicy, MatType>::BatchCluster(
   std::vector<std::vector<ElemType>> distances;
   Log::Info << "Performing range search." << std::endl;
   rangeSearch.Train(data);
-  rangeSearch.Search(RangeType<ElemType>(zero, epsilon), neighbors, distances);
+  rangeSearch.Search(RangeType<ElemType>(ElemType(0.0), epsilon), neighbors, distances);
   Log::Info << "Range search complete." << std::endl;
 
   // See the description of the algorithm in `PointwiseCluster()`.  The strategy

src/mlpack/methods/range_search/range_search.hpp

@@ -46,7 +46,8 @@ class RangeSearch
  public:
   //! Convenience typedef.
   typedef TreeType<MetricType, RangeSearchStat, MatType> Tree;
-
+   //! The type of Matrix.
+  typedef MatType Mat;
   //! The type of element held in MatType.
   typedef typename MatType::elem_type ElemType;
 

src/mlpack/tests/dbscan_test.cpp

@@ -230,8 +230,7 @@ TEST_CASE("Float32OutlierSingleModeTest", "[DBSCANTest]")
          EuclideanDistance,
          arma::Mat<float>,
          FloatKDTree>,
-         OrderedPointSelection,
-         arma::Mat<float>> d(0.1, 3, false);
+         OrderedPointSelection> d(0.1, 3, false);
 
   arma::Row<size_t> assignments;
   const size_t clusters = d.Cluster(points, assignments);