public void initAndValidate() { paramList = paramListInput.get(); modelList = modelListInput.get(); freqsList = freqsListInput.get(); ratesList = ratesListInput.get(); paramPointers = paramPointersInput.get(); modelPointers = modelPointersInput.get(); freqsPointers = freqsPointersInput.get(); ratesPointers = ratesPointersInput.get(); pointerCount = paramPointers.getDimension(); dp = dpInput.get(); List<ParametricDistribution> distrs = dp.getBaseDistributions(); paramBaseDistr = paramDistrInput.get(); // paramBaseDistr = distrs.get(0); modelBaseDistr = distrs.get(1); // modelBaseDistr = modelDistrInput.get(); freqsBaseDistr = distrs.get(2); // ratesBaseDistr = distrs.get(3); ratesBaseDistr = rateDistrInput.get(); tempLikelihood = tempLikelihoodInput.get(); dpTreeLikelihood = dpTreeLikelihoodInput.get(); modelNetworkMap.put(1.0, new double[] {3.0}); modelNetworkMap.put(2.0, new double[] {3.0}); modelNetworkMap.put(3.0, new double[] {1.0, 2.0, 4.0}); modelNetworkMap.put(4.0, new double[] {3.0, 5.0}); modelNetworkMap.put(5.0, new double[] {4.0}); // System.out.println("is null? "+(modelNetworkMap.get(5.0) == null)); }
public void initAndValidate() { testCorrect = testCorrectInput.get(); ratesList = ratesListInput.get(); alphaList = alphaListInput.get(); invPrList = invPrListInput.get(); siteModelList = siteModelListInput.get(); ratesPointers = ratesPointersInput.get(); pointerCount = ratesPointers.getDimension(); ratesBaseDistr = ratesDistrInput.get(); alphaBaseDistr = alphaDistrInput.get(); invPrBaseDistr = invPrDistrInput.get(); siteModelBaseDistr = siteModelDistrInput.get(); tempLikelihood = tempLikelihoodInput.get(); dpTreeLikelihood = dpTreeLikelihoodInput.get(); modelNetworkMap.put(1.0, new double[] {3.0}); modelNetworkMap.put(2.0, new double[] {3.0}); modelNetworkMap.put(3.0, new double[] {1.0, 2.0, 4.0}); modelNetworkMap.put(4.0, new double[] {3.0, 5.0}); modelNetworkMap.put(5.0, new double[] {4.0}); // System.out.println("is null? "+(modelNetworkMap.get(5.0) == null)); }
public double proposal() { double logq = 0.0; // Pick two indcies at random int index1 = Randomizer.nextInt(pointerCount); int index2 = index1; while (index2 == index1) { index2 = Randomizer.nextInt(pointerCount); } int clusterIndex1 = paramPointers.indexInList(index1, paramList); int clusterIndex2 = paramPointers.indexInList(index2, paramList); // If the randomly draw sites are from the same cluster, perform a split-move. if (clusterIndex1 == clusterIndex2) { int[] clusterSites = dpValuableInput.get().getClusterSites(clusterIndex1); double temp = split(index1, index2, clusterIndex1, clusterSites); // System.out.println("split: "+temp); logq += temp; // System.out.println("split: "+temp); } else { // If the the two randomly drawn sites are not from the same cluster, perform a merge-move. int[] cluster1Sites = dpValuableInput.get().getClusterSites(clusterIndex1); int[] cluster2Sites = dpValuableInput.get().getClusterSites(clusterIndex2); // logq = merge(index1, index2,clusterIndex1,clusterIndex2,cluster1Sites,cluster2Sites); double temp = merge(index1, index2, clusterIndex1, clusterIndex2, cluster1Sites, cluster2Sites); // System.out.println("merge: "+temp); logq = temp; } return logq; }
public double merge( int index1, int index2, int clusterIndex1, int clusterIndex2, int[] cluster1Sites, int[] cluster2Sites) { double logqMerge = 0.0; HashMap<Integer, Integer> siteMap = new HashMap<Integer, Integer>(); // The value of the merged cluster will have that of cluster 2 before the merge. QuietRealParameter mergedParam = paramList.getParameter(clusterIndex2); QuietRealParameter mergedModel = modelList.getParameter(clusterIndex2); QuietRealParameter mergedFreqs = freqsList.getParameter(clusterIndex2); QuietRealParameter mergedRates = ratesList.getParameter(clusterIndex2); // Create a vector that combines the site indices of the two clusters int[] mergedClusterSites = new int[cluster1Sites.length + cluster2Sites.length - 2]; int k = 0; for (int i = 0; i < cluster1Sites.length; i++) { // Point every member in cluster 1 to cluster 2 // paramPointers.point(cluster1Sites[i],mergedParam); // modelPointers.point(cluster1Sites[i],mergedModel); // freqsPointers.point(cluster1Sites[i],mergedFreqs); // ratesPointers.point(cluster1Sites[i],mergedRates); if (cluster1Sites[i] != index1) { // For all members that are not index 1, // record the cluster in which they have been before the merge, // and assign them to the combined vector. siteMap.put(cluster1Sites[i], clusterIndex1); mergedClusterSites[k++] = cluster1Sites[i]; } } for (int i = 0; i < cluster2Sites.length; i++) { // All members in cluster 2 remains in cluster2 so no new pointer assignments if (cluster2Sites[i] != index2) { // For all members that are not index 2, // record the cluster in which they have been before the merge, // and assign them to the combined vector. siteMap.put(cluster2Sites[i], clusterIndex2); mergedClusterSites[k++] = cluster2Sites[i]; } } try { // Create a weight vector of patterns to inform the temporary tree likelihood // which set of pattern likelihoods are to be computed. // int[] tempWeights = dpTreeLikelihood.getClusterWeights(clusterIndex1); /*int[] tempWeights = new int[tempLikelihood.m_data.get().getPatternCount()]; for(int i = 0; i < cluster1Sites.length; i++){ int patIndex = tempLikelihood.m_data.get().getPatternIndex(cluster1Sites[i]); tempWeights[patIndex] = 1; } tempLikelihood.setPatternWeights(tempWeights); double[] cluster1SitesCluster2ParamLogLik = tempLikelihood.calculateLogP( mergedParam, mergedModel, mergedFreqs, mergedRates, cluster1Sites, index1 ); */ k = 0; int[] sCluster1Sites = new int[cluster1Sites.length - 1]; for (int i = 0; i < cluster1Sites.length; i++) { if (cluster1Sites[i] != index1) { sCluster1Sites[k++] = cluster1Sites[i]; } } tempLikelihood.setupPatternWeightsFromSites(sCluster1Sites); double[] cluster1SitesCluster2ParamLogLik = tempLikelihood.calculateLogP( mergedParam, mergedModel, mergedFreqs, mergedRates, sCluster1Sites); // tempWeights = dpTreeLikelihood.getClusterWeights(clusterIndex2); /*tempWeights = new int[tempLikelihood.m_data.get().getPatternCount()]; for(int i = 0; i < cluster2Sites.length; i++){ int patIndex = tempLikelihood.m_data.get().getPatternIndex(cluster2Sites[i]); tempWeights[patIndex] = 1; } tempLikelihood.setPatternWeights(tempWeights); RealParameter removedParam = paramList.getParameter(clusterIndex1); RealParameter removedModel = modelList.getParameter(clusterIndex1); RealParameter removedFreqs = freqsList.getParameter(clusterIndex1); RealParameter removedRates = ratesList.getParameter(clusterIndex1); double[] cluster2SitesCluster1ParamLogLik = tempLikelihood.calculateLogP( removedParam, removedModel, removedFreqs, removedRates, cluster2Sites, index2 );*/ k = 0; int[] sCluster2Sites = new int[cluster2Sites.length - 1]; for (int i = 0; i < cluster2Sites.length; i++) { if (cluster2Sites[i] != index2) { sCluster2Sites[k++] = cluster2Sites[i]; } } tempLikelihood.setupPatternWeightsFromSites(sCluster2Sites); RealParameter removedParam = paramList.getParameter(clusterIndex1); RealParameter removedModel = modelList.getParameter(clusterIndex1); RealParameter removedFreqs = freqsList.getParameter(clusterIndex1); RealParameter removedRates = ratesList.getParameter(clusterIndex1); double[] cluster2SitesCluster1ParamLogLik = tempLikelihood.calculateLogP( removedParam, removedModel, removedFreqs, removedRates, sCluster2Sites); // System.out.println("populate logLik1:"); double[] logLik1 = new double[mergedClusterSites.length]; for (int i = 0; i < (cluster1Sites.length - 1); i++) { // System.out.println(clusterIndex1+" "+mergedClusterSites[i]); logLik1[i] = dpTreeLikelihood.getSiteLogLikelihood(clusterIndex1, mergedClusterSites[i]); } System.arraycopy( cluster2SitesCluster1ParamLogLik, 0, logLik1, cluster1Sites.length - 1, cluster2SitesCluster1ParamLogLik.length); double[] logLik2 = new double[mergedClusterSites.length]; System.arraycopy( cluster1SitesCluster2ParamLogLik, 0, logLik2, 0, cluster1SitesCluster2ParamLogLik.length); // System.out.println("populate logLik2:"); for (int i = cluster1SitesCluster2ParamLogLik.length; i < logLik2.length; i++) { // System.out.println(clusterIndex2+" // "+mergedClusterSites[i-cluster1SitesCluster2ParamLogLik.length]); logLik2[i] = dpTreeLikelihood.getSiteLogLikelihood(clusterIndex2, mergedClusterSites[i]); } double[] lik1 = new double[logLik1.length]; double[] lik2 = new double[logLik2.length]; // scale it so it may be more accuate double minLog; /*for(int i = 0; i < logLik1.length; i++){ minLog = Math.min(logLik1[i],logLik2[i]); if(minLog == logLik1[i]){ lik1[i] = 1.0; lik2[i] = Math.exp(logLik2[i] - minLog); }else{ lik1[i] = Math.exp(logLik1[i] - minLog); lik2[i] = 1.0; } }*/ for (int i = 0; i < logLik1.length; i++) { lik1[i] = Math.exp(logLik1[i]); lik2[i] = Math.exp(logLik2[i]); // System.out.println(lik1[i]+" "+lik2[i]); } // Create a set of indices for random permutation int[] shuffle = new int[mergedClusterSites.length]; for (int i = 0; i < shuffle.length; i++) { shuffle[i] = i; } Randomizer.shuffle(shuffle); int cluster1Count = 1; int cluster2Count = 1; int cluster; double psi1, psi2, cluster1Prob; for (int i = 0; i < mergedClusterSites.length; i++) { cluster = siteMap.get(mergedClusterSites[shuffle[i]]); psi1 = cluster1Count * lik1[shuffle[i]]; psi2 = cluster2Count * lik2[shuffle[i]]; /*testCorrectness(i,cluster, clusterIndex1,clusterIndex2,shuffle, mergedClusterSites, lik1,lik2);*/ cluster1Prob = psi1 / (psi1 + psi2); // System.out.println(cluster1Prob); if (cluster == clusterIndex1) { logqMerge += Math.log(cluster1Prob); cluster1Count++; } else if (cluster == clusterIndex2) { logqMerge += Math.log(1 - cluster1Prob); cluster2Count++; } else { throw new RuntimeException("Something is wrong."); } } logqMerge += // paramBaseDistr.calcLogP(removedParam) mergeValue(removedParam, mergedParam, paramBaseDistr) // + modelBaseDistr.calcLogP(removedModel) + mergeDiscreteValue(removedModel, mergedModel, modelDistrInput.get()) + freqsBaseDistr.calcLogP(removedFreqs) // + ratesBaseDistr.calcLogP(removedRates); + mergeValueInLogSpace(removedRates, mergedRates, ratesBaseDistr); if (logqMerge > Double.NEGATIVE_INFINITY) { paramList.mergeParameter(clusterIndex1, clusterIndex2); modelList.mergeParameter(clusterIndex1, clusterIndex2); freqsList.mergeParameter(clusterIndex1, clusterIndex2); ratesList.mergeParameter(clusterIndex1, clusterIndex2); for (int i = 0; i < cluster1Sites.length; i++) { // Point every member in cluster 1 to cluster 2 paramPointers.point(cluster1Sites[i], mergedParam); modelPointers.point(cluster1Sites[i], mergedModel); freqsPointers.point(cluster1Sites[i], mergedFreqs); ratesPointers.point(cluster1Sites[i], mergedRates); } } } catch (Exception e) { throw new RuntimeException(e); } return logqMerge; }
public double split(int index1, int index2, int clusterIndex, int[] initClusterSites) { try { double logqSplit = 0.0; // Create a parameter by sampling from the prior // QuietRealParameter newParam = getSample(paramBaseDistr, paramList.getUpper(), // paramList.getLower()); QuietRealParameter newParam = new QuietRealParameter(new Double[5]); // logqSplit += proposeNewValue(newParam, paramBaseDistr, paramList.getUpper(), // paramList.getLower()); double[] oldParamValues = new double[5]; for (int i = 0; i < oldParamValues.length; i++) { oldParamValues[i] = paramList.getValue(clusterIndex, i); } logqSplit += proposeNewValue2( newParam, oldParamValues, paramBaseDistr, paramList.getUpper(), paramList.getLower()); // QuietRealParameter newModel = getSample(modelBaseDistr, modelList.getUpper(), // modelList.getLower()); QuietRealParameter newModel = new QuietRealParameter(new Double[1]); logqSplit += proposeDiscreteValue( newModel, modelList.getValue(clusterIndex, 0), modelDistrInput.get(), modelList.getUpper(), modelList.getLower()); QuietRealParameter newFreqs = getSample(freqsBaseDistr, freqsList.getUpper(), freqsList.getLower()); // QuietRealParameter newRates = getSample(ratesBaseDistr, ratesList.getUpper(), // ratesList.getLower()); QuietRealParameter newRates = new QuietRealParameter(new Double[1]); logqSplit += proposalValueInLogSpace( newRates, ratesList.getValue(clusterIndex, 0), ratesBaseDistr, ratesList.getUpper(), ratesList.getLower()); // Remove the index 1 and index 2 from the cluster int[] clusterSites = new int[initClusterSites.length - 2]; int k = 0; for (int i = 0; i < initClusterSites.length; i++) { if (initClusterSites[i] != index1 && initClusterSites[i] != index2) { clusterSites[k++] = initClusterSites[i]; } } // Form a new cluster with index 1 paramPointers.point(index1, newParam); modelPointers.point(index1, newModel); freqsPointers.point(index1, newFreqs); ratesPointers.point(index1, newRates); // Shuffle the cluster_-{index_1,index_2} to obtain a random permutation Randomizer.shuffle(clusterSites); // Create the weight vector of site patterns according to the order of the shuffled index. /*int[] tempWeights = new int[tempLikelihood.m_data.get().getPatternCount()]; int patIndex; for(int i = 0; i < clusterSites.length; i++){ patIndex = tempLikelihood.m_data.get().getPatternIndex(clusterSites[i]); tempWeights[patIndex] = 1; } tempLikelihood.setPatternWeights(tempWeights);*/ tempLikelihood.setupPatternWeightsFromSites(clusterSites); // Site log likelihoods in the order of the shuffled sites double[] logLik1 = tempLikelihood.calculateLogP(newParam, newModel, newFreqs, newRates, clusterSites); double[] logLik2 = new double[clusterSites.length]; for (int i = 0; i < logLik2.length; i++) { logLik2[i] = dpTreeLikelihood.getSiteLogLikelihood(clusterIndex, clusterSites[i]); } double[] lik1 = new double[logLik1.length]; double[] lik2 = new double[logLik2.length]; double minLog; // scale it so it may be more accurate /*for(int i = 0; i < logLik1.length; i++){ minLog = Math.min(logLik1[i],logLik2[i]); if(minLog == logLik1[i]){ lik1[i] = 1.0; lik2[i] = Math.exp(logLik2[i] - minLog); }else{ lik1[i] = Math.exp(logLik1[i] - minLog); lik2[i] = 1.0; } }*/ for (int i = 0; i < logLik1.length; i++) { lik1[i] = Math.exp(logLik1[i]); lik2[i] = Math.exp(logLik2[i]); // System.out.println(lik1[i]+" "+lik2[i]); } /*for(int i = 0; i < clusterSites.length;i++){ System.out.println("clusterSites: "+clusterSites[i]); } System.out.println("index 1: "+index1+" index2: "+index2);*/ int cluster1Count = 1; int cluster2Count = 1; int[] sitesInCluster1 = new int[initClusterSites.length]; sitesInCluster1[0] = index1; // Assign members of the existing cluster (except for indice 1 and 2) randomly // to the existing and the new cluster double psi1, psi2, newClusterProb, draw; for (int i = 0; i < clusterSites.length; i++) { psi1 = cluster1Count * lik1[i]; psi2 = cluster2Count * lik2[i]; newClusterProb = psi1 / (psi1 + psi2); draw = Randomizer.nextDouble(); if (draw < newClusterProb) { // System.out.println("in new cluster: "+clusterSites[i]); sitesInCluster1[cluster1Count] = clusterSites[i]; // paramPointers.point(clusterSites[i],newParam); // modelPointers.point(clusterSites[i],newModel); // freqsPointers.point(clusterSites[i],newFreqs); // ratesPointers.point(clusterSites[i],newRates); logqSplit += Math.log(newClusterProb); cluster1Count++; } else { logqSplit += Math.log(1.0 - newClusterProb); cluster2Count++; } } // logqSplit += paramBaseDistr.calcLogP(newParam) logqSplit += // modelBaseDistr.calcLogP(newModel) + freqsBaseDistr.calcLogP(newFreqs) // + ratesBaseDistr.calcLogP(newRates) ; // Perform a split paramList = paramListInput.get(this); modelList = modelListInput.get(this); freqsList = freqsListInput.get(this); ratesList = ratesListInput.get(this); paramPointers = paramPointersInput.get(this); modelPointers = modelPointersInput.get(this); freqsPointers = freqsPointersInput.get(this); ratesPointers = ratesPointersInput.get(this); paramList.splitParameter(clusterIndex, newParam); modelList.splitParameter(clusterIndex, newModel); freqsList.splitParameter(clusterIndex, newFreqs); ratesList.splitParameter(clusterIndex, newRates); // Form a new cluster with index 1 paramPointers = paramPointersInput.get(this); modelPointers = modelPointersInput.get(this); freqsPointers = freqsPointersInput.get(this); ratesPointers = ratesPointersInput.get(this); for (int i = 0; i < cluster1Count; i++) { paramPointers.point(sitesInCluster1[i], newParam); modelPointers.point(sitesInCluster1[i], newModel); freqsPointers.point(sitesInCluster1[i], newFreqs); ratesPointers.point(sitesInCluster1[i], newRates); } return -logqSplit; } catch (Exception e) { throw new RuntimeException(e); } }
public double split(int index1, int index2, int clusterIndex, int[] initClusterSites) { try { double logqSplit = 0.0; // Create a parameter by sampling from the prior QuietRealParameter newParam = getSample(paramBaseDistr, paramList.getUpper(), paramList.getLower()); QuietRealParameter newModel = getSample(modelBaseDistr, modelList.getUpper(), modelList.getLower()); QuietRealParameter newFreqs = getSample(freqsBaseDistr, freqsList.getUpper(), freqsList.getLower()); // Perform a split // paramList.splitParameter(clusterIndex,newParam); // modelList.splitParameter(clusterIndex,newModel); // freqsList.splitParameter(clusterIndex,newFreqs); // Remove the index 1 and index 2 from the cluster int[] clusterSites = new int[initClusterSites.length - 2]; int k = 0; for (int i = 0; i < initClusterSites.length; i++) { if (initClusterSites[i] != index1 && initClusterSites[i] != index2) { clusterSites[k++] = initClusterSites[i]; } } // Form a new cluster with index 1 // paramPointers.point(index1,newParam); // modelPointers.point(index1,newModel); // freqsPointers.point(index1,newFreqs); // Shuffle the cluster_-{index_1,index_2} to obtain a random permutation Randomizer.shuffle(clusterSites); // Create the weight vector of site patterns according to the order of the shuffled index. /*int[] tempWeights = new int[tempLikelihood.m_data.get().getPatternCount()]; int patIndex; for(int i = 0; i < clusterSites.length; i++){ patIndex = tempLikelihood.m_data.get().getPatternIndex(clusterSites[i]); tempWeights[patIndex] = 1; }*/ tempLikelihood.setupPatternWeightsFromSites(clusterSites); // Site log likelihoods in the order of the shuffled sites double[] logLik1 = tempLikelihood.calculateLogP(newParam, newModel, newFreqs, clusterSites); double[] logLik2 = new double[clusterSites.length]; for (int i = 0; i < logLik2.length; i++) { // logLik2[i] = dpTreeLikelihood.getSiteLogLikelihood(clusterIndex,clusterSites[i]); logLik2[i] = getSiteLogLikelihood( paramList.getParameter(clusterIndex).getIDNumber(), clusterIndex, clusterSites[i]); } double[] lik1 = new double[logLik1.length]; double[] lik2 = new double[logLik2.length]; double maxLog; // scale it so it may be more accurate for (int i = 0; i < logLik1.length; i++) { maxLog = Math.max(logLik1[i], logLik2[i]); if (Math.exp(maxLog) < 1e-100) { if (maxLog == logLik1[i]) { lik1[i] = 1.0; lik2[i] = Math.exp(logLik2[i] - maxLog); } else { lik1[i] = Math.exp(logLik1[i] - maxLog); lik2[i] = 1.0; } } else { lik1[i] = Math.exp(logLik1[i]); lik2[i] = Math.exp(logLik2[i]); } } /*boolean ohCrap = false; for(int i = 0; i < logLik1.length; i++){ if(Double.isNaN(logLik1[i])){ return Double.NEGATIVE_INFINITY; //ohCrap = true; //System.out.println("logLik1: "+logLik1); //logLik1[i] = Double.NEGATIVE_INFINITY; } if(Double.isNaN(logLik2[i])){ return Double.NEGATIVE_INFINITY; //ohCrap = true; //System.out.println("logLik1: "+logLik2); //logLik2[i] = Double.NEGATIVE_INFINITY; } lik1[i] = Math.exp(logLik1[i]); lik2[i] = Math.exp(logLik2[i]); //System.out.println(lik1[i]+" "+lik2[i]); } if(ohCrap){ for(int i = 0; i < newParam.getDimension();i++){ System.out.print(newParam.getValue(i)+" "); } System.out.println(); }*/ /*for(int i = 0; i < clusterSites.length;i++){ System.out.println("clusterSites: "+clusterSites[i]); } System.out.println("index 1: "+index1+" index2: "+index2);*/ int cluster1Count = 1; int cluster2Count = 1; // Assign members of the existing cluster (except for indice 1 and 2) randomly // to the existing and the new cluster double psi1, psi2, newClusterProb, draw; int[] newAssignment = new int[clusterSites.length]; for (int i = 0; i < clusterSites.length; i++) { psi1 = cluster1Count * lik1[i]; psi2 = cluster2Count * lik2[i]; newClusterProb = psi1 / (psi1 + psi2); draw = Randomizer.nextDouble(); if (draw < newClusterProb) { // System.out.println("in new cluster: "+clusterSites[i]); // paramPointers.point(clusterSites[i],newParam); // modelPointers.point(clusterSites[i],newModel); // freqsPointers.point(clusterSites[i],newFreqs); newAssignment[cluster1Count - 1] = clusterSites[i]; logqSplit += Math.log(newClusterProb); cluster1Count++; } else { logqSplit += Math.log(1.0 - newClusterProb); cluster2Count++; } } // System.out.println("halfway: "+logqSplit); logqSplit += paramBaseDistr.calcLogP(newParam) + modelBaseDistr.calcLogP(newModel) + freqsBaseDistr.calcLogP(newFreqs); if (-logqSplit > Double.NEGATIVE_INFINITY) { paramList = paramListInput.get(this); modelList = modelListInput.get(this); freqsList = freqsListInput.get(this); paramPointers = paramPointersInput.get(this); modelPointers = modelPointersInput.get(this); freqsPointers = freqsPointersInput.get(this); // Perform a split paramList.splitParameter(clusterIndex, newParam); modelList.splitParameter(clusterIndex, newModel); freqsList.splitParameter(clusterIndex, newFreqs); // Form a new cluster with index 1 paramPointers.point(index1, newParam); modelPointers.point(index1, newModel); freqsPointers.point(index1, newFreqs); for (int i = 0; i < (cluster1Count - 1); i++) { paramPointers.point(newAssignment[i], newParam); modelPointers.point(newAssignment[i], newModel); freqsPointers.point(newAssignment[i], newFreqs); } } return -logqSplit; } catch (Exception e) { // freqsBaseDistr.printDetails(); throw new RuntimeException(e); } }