Source Code for Module rdkit.ML.InfoTheory.BitClusterer

 1  # 
 2  #  Copyright (C) 2000-2008  Greg Landrum and Rational Discovery LLC 
 3  # 
 4   
 5  from rdkit.SimDivFilters import rdSimDivPickers as rdsimdiv 
 6  from rdkit import DataStructs 
 7  import numpy 
 8   
 9 -class BitClusterer(object): 
10      """ Class to cluster a set of bits based on their correllation 
11   
12      The correlation matrix is first built using by reading the fingerprints 
13      from a database or a list of fingerprints 
14      """ 
15   
16 -    def __init__(self, idList, nCluster, type=rdsimdiv.ClusterMethod.WARD): 
17          self._clusters = [] 
18          self._bidList = idList 
19          #self._matGen = BitCorrelationMatGenerator(idList) 
20          self._nClusters = nCluster 
21          self._type = type 
22   
23 -    def ClusterBits(self, corrMat) : 
24          # clutering code actually needs distances so, take 1/val for each element in corMat 
25          distMat = 1/corrMat 
26   
27          pkr = rdsimdiv.HierarchicalClusterPicker(self._type) 
28           
29          cls = pkr.Cluster(distMat, len(self._bidList), self._nClusters) 
30          # map the clusters to the actual bit ids 
31          self._clusters = [] 
32          for cl in cls : 
33              bcls = [] 
34              for i in cl : 
35                  bid = self._bidList[i] 
36                  bcls.append(bid) 
37              self._clusters.append(bcls) 
38   
39 -    def SetClusters(self, clusters): 
40          assert len(clusters) == self._nClusters 
41          self._clusters = clusters 
42           
43 -    def GetClusters(self) : 
44          return self._clusters 
45   
46 -    def MapToClusterScores(self, fp) : 
47          """ Map the fingerprint to a real valued vector of score based on the bit clusters 
48   
49          The dimension of the vector is same as the number of clusters. Each value in the  
50          vector corresponds to the number of bits in the corresponding cluster 
51          that are turned on in the fingerprint 
52   
53          ARGUMENTS: 
54           - fp : the fingerprint  
55          """ 
56       
57          scores = [0]*self._nClusters 
58   
59          i = 0 
60          for cls in self._clusters: 
61              for bid in cls : 
62                  if fp[bid] : 
63                      scores[i] += 1 
64   
65              i += 1 
66   
67          return scores 
68   
69 -    def MapToClusterFP(self, fp) : 
70          """ Map the fingerprint to a smaller sized (= number of clusters) fingerprint 
71   
72          Each cluster get a bit in the new fingerprint and is turned on if any of the bits in 
73          the cluster are turned on in the original fingerprint""" 
74   
75          ebv = DataStructs.ExplicitBitVect(self._nClusters) 
76          i = 0 
77   
78          for cls in self._clusters: 
79              for bid in cls : 
80                  if fp[bid] : 
81                      ebv.SetBit(i) 
82                      break 
83              i += 1 
84           
85          return ebv 
86