Line data    Source code

       1             : /* +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
       2             :    Copyright (c) 2014-2020 The plumed team
       3             :    (see the PEOPLE file at the root of the distribution for a list of names)
       4             : 
       5             :    See http://www.plumed.org for more information.
       6             : 
       7             :    This file is part of plumed, version 2.
       8             : 
       9             :    plumed is free software: you can redistribute it and/or modify
      10             :    it under the terms of the GNU Lesser General Public License as published by
      11             :    the Free Software Foundation, either version 3 of the License, or
      12             :    (at your option) any later version.
      13             : 
      14             :    plumed is distributed in the hope that it will be useful,
      15             :    but WITHOUT ANY WARRANTY; without even the implied warranty of
      16             :    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
      17             :    GNU Lesser General Public License for more details.
      18             : 
      19             :    You should have received a copy of the GNU Lesser General Public License
      20             :    along with plumed.  If not, see <http://www.gnu.org/licenses/>.
      21             : +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ */
      22             : #include "ClusteringBase.h"
      23             : #include "core/ActionRegister.h"
      24             : 
      25             : #ifdef __PLUMED_HAS_BOOST_GRAPH
      26             : #include <boost/graph/adjacency_list.hpp>
      27             : #include <boost/graph/connected_components.hpp>
      28             : #include <boost/graph/graph_utility.hpp>
      29             : #endif
      30             : 
      31             : //+PLUMEDOC MATRIXF DFSCLUSTERING
      32             : /*
      33             : Find the connected components of the matrix using the depth first search clustering algorithm.
      34             : 
      35             : As discussed in the section of the manual on \ref contactmatrix a useful tool for developing complex collective variables is the notion of the
      36             : so called adjacency matrix.  An adjacency matrix is an \f$N \times N\f$ matrix in which the \f$i\f$th, \f$j\f$th element tells you whether
      37             : or not the \f$i\f$th and \f$j\f$th atoms/molecules from a set of \f$N\f$ atoms/molecules are adjacent or not.  As detailed in \cite tribello-clustering
      38             : these matrices provide a representation of a graph and can thus can be analyzed using tools from graph theory.  This particular action performs
      39             : a depth first search clustering to find the connected components of this graph.  You can read more about depth first search here:
      40             : 
      41             : https://en.wikipedia.org/wiki/Depth-first_search
      42             : 
      43             : This action is useful if you are looking at a phenomenon such as nucleation where the aim is to detect the sizes of the crystalline nuclei that have formed
      44             : in your simulation cell.
      45             : 
      46             : \par Examples
      47             : 
      48             : The input below calculates the coordination numbers of atoms 1-100 and then computes the an adjacency
      49             : matrix whose elements measures whether atoms \f$i\f$ and \f$j\f$ are within 0.55 nm of each other.  The action
      50             : labelled dfs then treats the elements of this matrix as zero or ones and thus thinks of the matrix as defining
      51             : a graph.  This dfs action then finds the largest connected component in this graph.  The sum of the coordination
      52             : numbers for the atoms in this largest connected component are then computed and this quantity is output to a colvar
      53             : file.  The way this input can be used is described in detail in \cite tribello-clustering.
      54             : 
      55             : \plumedfile
      56             : lq: COORDINATIONNUMBER SPECIES=1-100 SWITCH={CUBIC D_0=0.45  D_MAX=0.55} LOWMEM
      57             : cm: CONTACT_MATRIX ATOMS=lq  SWITCH={CUBIC D_0=0.45  D_MAX=0.55}
      58             : dfs: DFSCLUSTERING MATRIX=cm
      59             : clust1: CLUSTER_PROPERTIES CLUSTERS=dfs CLUSTER=1 SUM
      60             : PRINT ARG=clust1.* FILE=colvar
      61             : \endplumedfile
      62             : 
      63             : */
      64             : //+ENDPLUMEDOC
      65             : 
      66             : namespace PLMD {
      67             : namespace clusters {
      68             : 
      69             : class DFSClustering : public ClusteringBase {
      70             : private:
      71             : #ifndef __PLUMED_HAS_BOOST_GRAPH
      72             : /// The number of neighbors each atom has
      73             :   std::vector<unsigned> nneigh;
      74             : /// The adjacency list
      75             :   Matrix<unsigned> adj_list;
      76             : /// The color that tells us whether a node has been visited
      77             :   std::vector<unsigned> color;
      78             : /// The recursive function at the heart of this method
      79             :   int explore( const unsigned& index );
      80             : #endif
      81             : public:
      82             : /// Create manual
      83             :   static void registerKeywords( Keywords& keys );
      84             : /// Constructor
      85             :   explicit DFSClustering(const ActionOptions&);
      86             : /// Do the clustering
      87             :   void performClustering() override;
      88             : };
      89             : 
      90             : PLUMED_REGISTER_ACTION(DFSClustering,"DFSCLUSTERING")
      91             : 
      92          21 : void DFSClustering::registerKeywords( Keywords& keys ) {
      93          21 :   ClusteringBase::registerKeywords( keys );
      94          42 :   keys.addFlag("LOWMEM",false,"this flag does nothing and is present only to ensure back-compatibility");
      95          21 : }
      96             : 
      97          17 : DFSClustering::DFSClustering(const ActionOptions&ao):
      98             :   Action(ao),
      99          17 :   ClusteringBase(ao) {
     100             : #ifndef __PLUMED_HAS_BOOST_GRAPH
     101          17 :   nneigh.resize( getNumberOfNodes() );
     102          17 :   color.resize(getNumberOfNodes());
     103             : #endif
     104             :   bool lowmem;
     105          17 :   parseFlag("LOWMEM",lowmem);
     106          17 :   if( lowmem ) {
     107           0 :     warning("LOWMEM flag is deprecated and is no longer required for this action");
     108             :   }
     109          17 : }
     110             : 
     111          38 : void DFSClustering::performClustering() {
     112             : #ifdef __PLUMED_HAS_BOOST_GRAPH
     113             :   // Get the list of edges
     114             :   unsigned nedges=0;
     115             :   retrieveEdgeList( 0, nedges );
     116             : 
     117             :   // Build the graph using boost
     118             :   boost::adjacency_list<boost::vecS,boost::vecS,boost::undirectedS> sg(&pairs[0],&pairs[nedges],getNumberOfNodes());
     119             : 
     120             :   // Find the connected components using boost (-1 here for compatibility with non-boost version)
     121             :   number_of_cluster=boost::connected_components(sg,&which_cluster[0]) - 1;
     122             : 
     123             :   // And work out the size of each cluster
     124             :   for(unsigned i=0; i<which_cluster.size(); ++i) {
     125             :     cluster_sizes[which_cluster[i]].first++;
     126             :   }
     127             : #else
     128             :   // Get the adjacency matrix
     129          38 :   retrieveAdjacencyLists( nneigh, adj_list );
     130             : 
     131             :   // Perform clustering
     132          38 :   number_of_cluster=-1;
     133          38 :   color.assign(color.size(),0);
     134       20912 :   for(unsigned i=0; i<getNumberOfNodes(); ++i) {
     135       20874 :     if( color[i]==0 ) {
     136       11261 :       number_of_cluster++;
     137       11261 :       color[i]=explore(i);
     138             :     }
     139             :   }
     140             : #endif
     141          38 : }
     142             : 
     143             : #ifndef __PLUMED_HAS_BOOST_GRAPH
     144       20874 : int DFSClustering::explore( const unsigned& index ) {
     145             : 
     146       20874 :   color[index]=1;
     147      153802 :   for(unsigned i=0; i<nneigh[index]; ++i) {
     148      132928 :     unsigned j=adj_list(index,i);
     149      132928 :     if( color[j]==0 ) {
     150        9613 :       color[j]=explore(j);
     151             :     }
     152             :   }
     153             : 
     154             :   // Count the size of the cluster
     155       20874 :   cluster_sizes[number_of_cluster].first++;
     156       20874 :   which_cluster[index] = number_of_cluster;
     157       20874 :   return color[index];
     158             : }
     159             : #endif
     160             : 
     161             : }
     162             : }