Example to compute persistence of a Rips filtration "in parallel". On a machine

with a gazillion cores, it gains a factor >2 at the expense of some memory. At
this point, construction of the Rips graph and expansion are not negligible
anymore and could use some parallelism.


git-svn-id: svn+ssh://scm.gforge.inria.fr/svnroot/gudhi/branches/tbb@875 636b058d-ea47-450e-bf9e-a15bfbe3eedb

Former-commit-id: 53d6a74338f1346bd7d1ca018ec15dd1ee8287e8

2 files changed, 187 insertions, 7 deletions

diff --git a/src/Hasse_complex/include/gudhi/Hasse_complex.h b/src/Hasse_complex/include/gudhi/Hasse_complex.h
index af9ae5e9..38887264 100644
--- a/src/Hasse_complex/include/gudhi/Hasse_complex.h
+++ b/src/Hasse_complex/include/gudhi/Hasse_complex.h
@@ -44,8 +44,7 @@ struct Hasse_simplex {
   template< class Complex_ds >
   Hasse_simplex(Complex_ds & cpx
                 , typename Complex_ds::Simplex_handle sh)
-      : key_(cpx.key(sh))
-      , filtration_(cpx.filtration(sh))
+      : filtration_(cpx.filtration(sh))
       , boundary_() {
     boundary_.reserve(cpx.dimension(sh) + 1);
     for (auto b_sh : cpx.boundary_simplex_range(sh)) {
@@ -55,7 +54,7 @@ struct Hasse_simplex {
 
   Hasse_simplex(typename HasseCpx::Simplex_key key
                 , typename HasseCpx::Filtration_value fil
-                , std::vector<typename HasseCpx::Simplex_handle> boundary)
+                , std::vector<typename HasseCpx::Simplex_handle> const& boundary)
       : key_(key)
       , filtration_(fil)
       , boundary_(boundary) { }
@@ -197,11 +196,12 @@ class Hasse_complex {
   }
 
   void initialize_filtration() {
+    // Setting the keys is done by pcoh, Simplex_tree doesn't do it either.
+#if 0
     Simplex_key key = 0;
-    for (auto & h_simp : complex_) {
-      h_simp.key_ = key;
-      ++key;
-    }
+    for (auto & h_simp : complex_)
+      h_simp.key_ = key++;
+#endif
   }
 
   std::vector< Hasse_simp, Gudhi::no_init_allocator<Hasse_simp> > complex_;
diff --git a/src/Persistent_cohomology/example/parallel_rips_persistence.cpp b/src/Persistent_cohomology/example/parallel_rips_persistence.cpp
new file mode 100644
index 00000000..425a5b2c
--- /dev/null
+++ b/src/Persistent_cohomology/example/parallel_rips_persistence.cpp
@@ -0,0 +1,180 @@
+/*    This file is part of the Gudhi Library. The Gudhi library 
+ *    (Geometric Understanding in Higher Dimensions) is a generic C++ 
+ *    library for computational topology.
+ *
+ *    Author(s):       Clément Maria, Marc Glisse
+ *
+ *    Copyright (C) 2014  INRIA Sophia Antipolis-Méditerranée (France),
+ *                  2015  INRIA Saclay Île de France)
+ *
+ *    This program is free software: you can redistribute it and/or modify
+ *    it under the terms of the GNU General Public License as published by
+ *    the Free Software Foundation, either version 3 of the License, or
+ *    (at your option) any later version.
+ *
+ *    This program is distributed in the hope that it will be useful,
+ *    but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ *    GNU General Public License for more details.
+ *
+ *    You should have received a copy of the GNU General Public License
+ *    along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include <gudhi/reader_utils.h>
+#include <gudhi/graph_simplicial_complex.h>
+#include <gudhi/distance_functions.h>
+#include <gudhi/Simplex_tree.h>
+#include <gudhi/Persistent_cohomology.h>
+#include "gudhi/Hasse_complex.h"
+
+#include <boost/program_options.hpp>
+
+#ifdef GUDHI_USE_TBB
+#include <tbb/task_scheduler_init.h>
+#endif
+
+#include <string>
+#include <vector>
+
+////////////////////////////////////////////////////////////////
+//                                                            //
+//  WARNING: persistence computation itself is not parallel,  //
+//  and this uses more memory than rips_persistence.          //
+//                                                            //
+////////////////////////////////////////////////////////////////
+
+using namespace Gudhi;
+using namespace Gudhi::persistent_cohomology;
+
+typedef int Vertex_handle;
+typedef double Filtration_value;
+
+void program_options(int argc, char * argv[]
+                     , std::string & filepoints
+                     , std::string & filediag
+                     , Filtration_value & threshold
+                     , int & dim_max
+                     , int & p
+                     , Filtration_value & min_persistence);
+
+int main(int argc, char * argv[]) {
+  std::string filepoints;
+  std::string filediag;
+  Filtration_value threshold;
+  int dim_max;
+  int p;
+  Filtration_value min_persistence;
+
+  program_options(argc, argv, filepoints, filediag, threshold, dim_max, p, min_persistence);
+
+  // Extract the points from the file filepoints
+  typedef std::vector<double> Point_t;
+  std::vector< Point_t > points;
+  read_points(filepoints, points);
+
+  // Compute the proximity graph of the points
+  Graph_t prox_graph = compute_proximity_graph(points, threshold
+                                               , euclidean_distance<Point_t>);
+
+  // Construct the Rips complex in a Simplex Tree
+  Simplex_tree<>& st = *new Simplex_tree<>;
+  // insert the proximity graph in the simplex tree
+  st.insert_graph(prox_graph);
+  // expand the graph until dimension dim_max
+  st.expansion(dim_max);
+
+  std::cout << "The complex contains " << st.num_simplices() << " simplices \n";
+  std::cout << "   and has dimension " << st.dimension() << " \n";
+
+#ifdef GUDHI_USE_TBB
+  // Unnecessary, but clarifies which operations are parallel.
+  tbb::task_scheduler_init ts;
+#endif
+
+  // Sort the simplices in the order of the filtration
+  st.initialize_filtration();
+  int count = 0;
+  for (auto sh : st.filtration_simplex_range())
+    st.assign_key(sh, count++);
+
+  // Convert to a more convenient representation.
+  Hasse_complex<> hcpx(st);
+
+#ifdef GUDHI_USE_TBB
+  ts.terminate();
+#endif
+
+  // Free some space.
+  delete &st;
+
+  // Compute the persistence diagram of the complex
+  persistent_cohomology::Persistent_cohomology< Hasse_complex<>, Field_Zp > pcoh(hcpx);
+  // initializes the coefficient field for homology
+  pcoh.init_coefficients(p);
+
+  pcoh.compute_persistent_cohomology(min_persistence);
+
+  // Output the diagram in filediag
+  if (filediag.empty()) {
+    pcoh.output_diagram();
+  } else {
+    std::ofstream out(filediag);
+    pcoh.output_diagram(out);
+    out.close();
+  }
+}
+
+void program_options(int argc, char * argv[]
+                     , std::string & filepoints
+                     , std::string & filediag
+                     , Filtration_value & threshold
+                     , int & dim_max
+                     , int & p
+                     , Filtration_value & min_persistence) {
+  namespace po = boost::program_options;
+  po::options_description hidden("Hidden options");
+  hidden.add_options()
+      ("input-file", po::value<std::string>(&filepoints),
+       "Name of file containing a point set. Format is one point per line:   X1 ... Xd ");
+
+  po::options_description visible("Allowed options", 100);
+  visible.add_options()
+      ("help,h", "produce help message")
+      ("output-file,o", po::value<std::string>(&filediag)->default_value(std::string()),
+       "Name of file in which the persistence diagram is written. Default print in std::cout")
+      ("max-edge-length,r", po::value<Filtration_value>(&threshold)->default_value(0),
+       "Maximal length of an edge for the Rips complex construction.")
+      ("cpx-dimension,d", po::value<int>(&dim_max)->default_value(1),
+       "Maximal dimension of the Rips complex we want to compute.")
+      ("field-charac,p", po::value<int>(&p)->default_value(11),
+       "Characteristic p of the coefficient field Z/pZ for computing homology.")
+      ("min-persistence,m", po::value<Filtration_value>(&min_persistence),
+       "Minimal lifetime of homology feature to be recorded. Default is 0. Enter a negative value to see zero length intervals");
+
+  po::positional_options_description pos;
+  pos.add("input-file", 1);
+
+  po::options_description all;
+  all.add(visible).add(hidden);
+
+  po::variables_map vm;
+  po::store(po::command_line_parser(argc, argv).
+            options(all).positional(pos).run(), vm);
+  po::notify(vm);
+
+  if (vm.count("help") || !vm.count("input-file")) {
+    std::cout << std::endl;
+    std::cout << "Compute the persistent homology with coefficient field Z/pZ \n";
+    std::cout << "of a Rips complex defined on a set of input points.\n \n";
+    std::cout << "The output diagram contains one bar per line, written with the convention: \n";
+    std::cout << "   p   dim b d \n";
+    std::cout << "where dim is the dimension of the homological feature,\n";
+    std::cout << "b and d are respectively the birth and death of the feature and \n";
+    std::cout << "p is the characteristic of the field Z/pZ used for homology coefficients." << std::endl << std::endl;
+
+    std::cout << "Usage: " << argv[0] << " [options] input-file" << std::endl << std::endl;
+    std::cout << visible << std::endl;
+    std::abort();
+  }
+}