Merge branch 'fix_various_defects' into 'main'

Fix various defects

Closes #11, #10, #9, #8, #7, #6, and #5

See merge request zhangc/mcpevent2hist!17

sophiread/README.md

@@ -55,7 +55,12 @@ The following steps have been tested under MaxOS and Linux, and it is in theory
   - `sophiread-<version>-Linux.tar.Z`: a tarball for Linux with TZ compression
   - `sophiread-<version>-Linux.sh`: an installer for Linux
 
-> Mac users should use `environment_mac.yml` instead. Also install [MacTex](https://www.tug.org/mactex/) before building the documentation.
+- For Mac users with m-series chip, please make the following adjustment:
+  - Create env with `CONDA_SUBDIR=osx-64 conda create -f environment_mac.yml`
+    - Currently `mlpack` does not have a `arm64` package from conda, so we need to fallback to x86-64 using Rosetta 2 under the hood.
+  - Install [MacTex](https://www.tug.org/mactex/) before building the documentation.
+  - __DO NOT__ install `mlpack` from homebrew.
+    - `mlpack` from homebrew can lead to linking error when building the DBSCAN object.
 
 Use the CLI
 -----------

sophiread/include/abs.h

@@ -33,8 +33,8 @@ class ABS : public ClusteringAlgorithm {
   std::vector<int> clusterLabels_;   // The cluster labels for each hit
   std::vector<std::vector<int>> clusterIndices_;  // The cluster indices for
                                                   // each cluster
-  const int numClusters_ = 128;      // The number of clusters use in runtime
-  const int maxClusterSize_ = 10;    // The maximum cluster size
-  const int spiderTimeRange_ = 75;   // The spider time range (in ns)
+  const int numClusters_ = 128;     // The number of clusters use in runtime
+  const int maxClusterSize_ = 10;   // The maximum cluster size
+  const int spiderTimeRange_ = 75;  // The spider time range (in ns)
   PeakFittingAlgorithm* peakFittingAlgorithm_;  // The clustering algorithm
 };

sophiread/include/dbscan.h

@@ -3,41 +3,55 @@
 #include "clustering.h"
 
 class DBSCAN : public ClusteringAlgorithm {
-public:
-  DBSCAN(double eps_time, size_t min_points_time, double eps_xy, size_t min_points_xy)
-      : m_eps_time(eps_time), m_min_points_time(min_points_time), m_eps_xy(eps_xy), m_min_points_xy(min_points_xy){};
+ public:
+  DBSCAN(double eps_time, size_t min_points_time, double eps_xy,
+         size_t min_points_xy)
+      : m_eps_time(eps_time),
+        m_min_points_time(min_points_time),
+        m_eps_xy(eps_xy),
+        m_min_points_xy(min_points_xy){};
   ~DBSCAN() = default;
-public:
+
+ public:
   void set_method(std::string method) { m_method = method; };
   void reset();
   std::vector<int> get_cluster_labels();
-  void fit(const std::vector<Hit>& hits);
-  std::vector<NeutronEvent> get_events(const std::vector<Hit>& hits);
-private:
-  class TimeClusterInfo
-  {
-  public:
+  void fit(const std::vector<Hit> &hits);
+  std::vector<NeutronEvent> get_events(const std::vector<Hit> &hits);
+  bool verbose() const { return m_verbose; }
+  void set_verbose(bool verbose) { m_verbose = verbose; }
+
+ private:
+  class TimeClusterInfo {
+   public:
     TimeClusterInfo();
     TimeClusterInfo(const double time, const size_t xy_index);
     ~TimeClusterInfo() = default;
-  public:
+
+   public:
     double m_time_mean;
     double m_time_sum;
     double m_time_min;
     double m_time_max;
-    std::vector<size_t> m_time_cluster_xy_indexes; // wrt input hits vector
+    std::vector<size_t> m_time_cluster_xy_indexes;  // wrt input hits vector
   };
-  void fit1D(std::vector<double> &data, size_t &number_of_clusters, std::vector<size_t> &labels, std::vector<double> &centroids);
-  void fit2D(std::vector<std::pair<double,double>> &data, size_t &number_of_clusters, std::vector<size_t> &labels, 
-      std::vector<std::pair<double,double>> &centroids);
-  void mergeTimeClusters1D(std::vector<TimeClusterInfo>& input_infos, std::vector<TimeClusterInfo>& merged_infos);
-private:
+  void fit1D(std::vector<double> &data, size_t &number_of_clusters,
+             std::vector<size_t> &labels, std::vector<double> &centroids);
+  void fit2D(std::vector<std::pair<double, double>> &data,
+             size_t &number_of_clusters, std::vector<size_t> &labels,
+             std::vector<std::pair<double, double>> &centroids);
+  void mergeTimeClusters1D(std::vector<TimeClusterInfo> &input_infos,
+                           std::vector<TimeClusterInfo> &merged_infos);
+
+ private:
   std::string m_method{"centroid"};  // method for centroid
-  double m_eps_time;          // The maximum distance between two time points
-  size_t m_min_points_time;   // The minimum number of points in a time cluster
-  double m_eps_xy;            // The maximum distance between two XY points
-  size_t m_min_points_xy;     // The minimum number of points in an XY cluster
+  double m_eps_time;         // The maximum distance between two time points
+  size_t m_min_points_time;  // The minimum number of points in a time cluster
+  double m_eps_xy;           // The maximum distance between two XY points
+  size_t m_min_points_xy;    // The minimum number of points in an XY cluster
   std::vector<NeutronEvent> m_events;
   const size_t m_max_hit_chunk_size = 2e6;
+  std::vector<int>
+      clusterLabels_;  // The cluster labels for each hit, not implemented yet
+  bool m_verbose = false;
 };
-

sophiread/src/centroid.cpp

@@ -1,7 +1,9 @@
 #include "centroid.h"
-#include "tpx3.h"
+
 #include <iostream>
 
+#include "tpx3.h"
+
 /**
  * @brief Perform centroid fitting on the hits.
  *
@@ -20,8 +22,8 @@ NeutronEvent Centroid::fit(const std::vector<Hit>& data) {
 
   if (weighted_by_tot) {
     for (const auto& hit : data) {
-      x += DSCALE*hit.getX() * hit.getTOT();
-      y += DSCALE*hit.getY() * hit.getTOT();
+      x += DSCALE * hit.getX() * hit.getTOT();
+      y += DSCALE * hit.getY() * hit.getTOT();
       tof += hit.getTOF();
       tot += hit.getTOT();
     }