protected double getRawBranchRate(final Tree tree, final NodeRef node) {
double rate = 0.0;
double[] processValues = getProcessValues(tree, node);
double totalTime = 0;
if (indicatorParameter != null) {
double absRate = rateParameter.getParameterValue(0);
for (int i = 0; i < indicatorParameter.getDimension(); i++) {
int index = (int) indicatorParameter.getParameterValue(i);
if (index == 0) {
rate += absRate * processValues[i];
} else {
rate +=
(absRate * (1.0 + relativeRatesParameter.getParameterValue(index - 1)))
* processValues[i];
}
totalTime += processValues[i];
}
} else {
for (int i = 0; i < rateParameter.getDimension(); i++) {
rate += rateParameter.getParameterValue(i) * processValues[i];
totalTime += processValues[i];
}
}
rate /= totalTime;
return rate;
}
/**
* @param name the name of the covarion substitution model
* @param dataType the data type
* @param freqModel the equlibrium frequencies
* @param hiddenClassRates the relative rates of the hidden categories (first hidden category has
* rate 1.0 so this parameter has dimension one less than number of hidden categories. each
* hidden category.
* @param switchingRates rate of switching between hidden categories
*/
public AbstractCovarionDNAModel(
String name,
HiddenNucleotides dataType,
Parameter hiddenClassRates,
Parameter switchingRates,
FrequencyModel freqModel) {
super(name, dataType, freqModel);
hiddenClassCount = dataType.getHiddenClassCount();
this.hiddenClassRates = hiddenClassRates;
this.switchingRates = switchingRates;
assert hiddenClassRates.getDimension() == hiddenClassCount - 1;
int hiddenClassCount = getHiddenClassCount();
int switchingClassCount = hiddenClassCount * (hiddenClassCount - 1) / 2;
if (switchingRates.getDimension() != switchingClassCount) {
throw new IllegalArgumentException(
"switching rate parameter must have "
+ switchingClassCount
+ " rates for "
+ hiddenClassCount
+ " classes");
}
addVariable(switchingRates);
addVariable(hiddenClassRates);
constructRateMatrixMap();
}
private void printInformtion(Parameter par) {
StringBuffer sb = new StringBuffer("parameter \n");
for (int j = 0; j < par.getDimension(); j++) {
sb.append(par.getParameterValue(j));
}
Logger.getLogger("dr.evomodel").info(sb.toString());
};
public Object parseXMLObject(XMLObject xo) throws XMLParseException {
final boolean onLogitScale = xo.getAttribute(ON_LOGIT_SCALE, false);
XMLObject cxo = xo.getChild(COUNTS);
Parameter countsParam = (Parameter) cxo.getChild(Parameter.class);
cxo = xo.getChild(PROPORTION);
Parameter proportionParam = (Parameter) cxo.getChild(Parameter.class);
if (proportionParam.getDimension() != 1
&& proportionParam.getDimension() != countsParam.getDimension()) {
throw new XMLParseException(
"Proportion dimension ("
+ proportionParam.getDimension()
+ ") "
+ "must equal 1 or counts dimension ("
+ countsParam.getDimension()
+ ")");
}
cxo = xo.getChild(TRIALS);
Parameter trialsParam;
if (cxo.hasAttribute(VALUES)) {
int[] tmp = cxo.getIntegerArrayAttribute(VALUES);
double[] v = new double[tmp.length];
for (int i = 0; i < tmp.length; ++i) {
v[i] = tmp[i];
}
trialsParam = new Parameter.Default(v);
} else {
trialsParam = (Parameter) cxo.getChild(Parameter.class);
}
if (trialsParam.getDimension() != countsParam.getDimension()) {
throw new XMLParseException(
"Trials dimension ("
+ trialsParam.getDimension()
+ ") must equal counts dimension ("
+ countsParam.getDimension()
+ ")");
}
return new BinomialLikelihood(trialsParam, proportionParam, countsParam, onLogitScale);
}
public MultivariateNormalIndependenceSampler(
Parameter parameter,
SelfControlledCaseSeries sccs,
double setSizeMean,
double weight,
double scaleFactor,
CoercionMode mode) {
super(mode);
this.scaleFactor = scaleFactor;
this.parameter = parameter;
setWeight(weight);
dim = parameter.getDimension();
setWeight(weight);
this.sccs = sccs;
this.setSizeMean = setSizeMean;
}
/** change the parameter and return the hastings ratio. */
public double doOperation() throws OperatorFailedException {
double[] mean = sccs.getMode();
double[] currentValue = parameter.getParameterValues();
double[] newValue = new double[dim];
Set<Integer> updateSet = new HashSet<Integer>();
if (setSizeMean != -1.0) {
final int listLength = Poisson.nextPoisson(setSizeMean);
while (updateSet.size() < listLength) {
int newInt = MathUtils.nextInt(parameter.getDimension());
if (!updateSet.contains(newInt)) {
updateSet.add(newInt);
}
}
} else {
for (int i = 0; i < dim; ++i) {
updateSet.add(i);
}
}
double logq = 0;
for (Integer i : updateSet) {
newValue[i] = mean[i] + scaleFactor * MathUtils.nextGaussian();
if (UPDATE_ALL) {
parameter.setParameterValueQuietly(i, newValue[i]);
} else {
parameter.setParameterValue(i, newValue[i]);
}
logq +=
(NormalDistribution.logPdf(currentValue[i], mean[i], scaleFactor)
- NormalDistribution.logPdf(newValue[i], mean[i], scaleFactor));
}
// for (Integer i : updateSet) {
// parameter.setParameterValueQuietly(i, newValue[i]);
// }
if (UPDATE_ALL) {
parameter.setParameterValueNotifyChangedAll(0, parameter.getParameterValue(0));
}
return logq;
}
public Object parseXMLObject(XMLObject xo) throws XMLParseException {
// final int overSampling = xo.getAttribute(OVERSAMPLING, 1);
final boolean normalize = xo.getAttribute(NORMALIZE, false);
final double normalizeBranchRateTo = xo.getAttribute(NORMALIZE_BRANCH_RATE_TO, Double.NaN);
TreeModel tree = (TreeModel) xo.getChild(TreeModel.class);
ParametricDistributionModel distributionModel =
(ParametricDistributionModel) xo.getElementFirstChild(DISTRIBUTION);
// Parameter rateCategoryParameter = (Parameter) xo.getElementFirstChild(RATE_CATEGORIES);
Parameter rateCategoryQuantilesParameter =
(Parameter) xo.getElementFirstChild(RATE_CATEGORY_QUANTILES);
Logger.getLogger("dr.evomodel").info("Using continuous relaxed clock model.");
// Logger.getLogger("dr.evomodel").info(" over sampling = " + overSampling);
Logger.getLogger("dr.evomodel")
.info(" parametric model = " + distributionModel.getModelName());
// Logger.getLogger("dr.evomodel").info(" rate categories = " +
// rateCategoryParameter.getDimension());
Logger.getLogger("dr.evomodel")
.info(" rate categories = " + rateCategoryQuantilesParameter.getDimension());
if (normalize) {
Logger.getLogger("dr.evomodel")
.info(" mean rate is normalized to " + normalizeBranchRateTo);
}
if (xo.hasAttribute(SINGLE_ROOT_RATE)) {
// singleRootRate = xo.getBooleanAttribute(SINGLE_ROOT_RATE);
Logger.getLogger("dr.evomodel").warning(" WARNING: single root rate is not implemented!");
}
/* if (xo.hasAttribute(NORMALIZED_MEAN)) {
dbr.setNormalizedMean(xo.getDoubleAttribute(NORMALIZED_MEAN));
}*/
return new ContinuousBranchRates(
tree, /*rateCategoryParameter, */
rateCategoryQuantilesParameter,
distributionModel, /*overSampling,*/
normalize,
normalizeBranchRateTo);
}
public int[] connected(int i, Parameter clusteringParameter) {
int n = clusteringParameter.getDimension();
int[] visited = new int[n + 1];
visited[0] = i + 1;
int tv = 1;
for (int j = 0; j < n; j++) {
if (visited[j] != 0) {
int curr = visited[j] - 1;
/*look forward
*/
int forward = (int) clusteringParameter.getParameterValue(curr);
visited[tv] = forward + 1;
tv++;
// Check to see if is isn't already on the list
for (int ii = 0; ii < tv - 1; ii++) {
if (visited[ii] == forward + 1) {
tv--;
visited[tv] = 0;
}
}
/*look back
*/
for (int jj = 0; jj < n; jj++) {
if ((int) clusteringParameter.getParameterValue(jj) == curr) {
visited[tv] = jj + 1;
tv++;
for (int ii = 0; ii < tv - 1; ii++) {
if (visited[ii] == jj + 1) {
tv--;
visited[tv] = 0;
}
}
}
}
}
}
return visited;
}
public DistanceDependentCRPGibbsOperator(
Parameter links,
Parameter assignments,
Parameter chiParameter,
NPAntigenicLikelihood Likelihood,
double weight) {
this.links = links;
this.assignments = assignments;
this.modelLikelihood = Likelihood;
this.chiParameter = chiParameter;
this.depMatrix = Likelihood.getLogDepMatrix();
for (int i = 0; i < links.getDimension(); i++) {
links.setParameterValue(i, i);
}
setWeight(weight);
// double[][] x=modelLikelihood.getData();
// modelLikelihood.printInformtion(x[0][0]);
this.m = new double[2];
m[0] = modelLikelihood.priorMean.getParameterValue(0);
m[1] = modelLikelihood.priorMean.getParameterValue(1);
this.v0 = 2;
this.k0 =
modelLikelihood.priorPrec.getParameterValue(0)
/ modelLikelihood.clusterPrec.getParameterValue(0);
this.T0Inv = new double[2][2];
T0Inv[0][0] = v0 / modelLikelihood.clusterPrec.getParameterValue(0);
T0Inv[1][1] = v0 / modelLikelihood.clusterPrec.getParameterValue(0);
T0Inv[1][0] = 0.0;
T0Inv[0][1] = 0.0;
this.logDetT0 = -Math.log(T0Inv[0][0] * T0Inv[1][1]);
}
public Object parseXMLObject(XMLObject xo) throws XMLParseException {
Parameter kappaParam;
Parameter switchingRates;
Parameter hiddenClassRates;
FrequencyModel freqModel;
kappaParam = (Parameter) xo.getElementFirstChild(KAPPA);
switchingRates = (Parameter) xo.getElementFirstChild(AbstractCovarionDNAModel.SWITCHING_RATES);
hiddenClassRates =
(Parameter) xo.getElementFirstChild(AbstractCovarionDNAModel.HIDDEN_CLASS_RATES);
freqModel = (FrequencyModel) xo.getElementFirstChild(AbstractCovarionDNAModel.FREQUENCIES);
if (!(freqModel.getDataType() instanceof HiddenNucleotides)) {
throw new IllegalArgumentException("Datatype must be hidden nucleotides!!");
}
HiddenNucleotides dataType = (HiddenNucleotides) freqModel.getDataType();
int hiddenStateCount = dataType.getHiddenClassCount();
int switchingRatesCount = hiddenStateCount * (hiddenStateCount - 1) / 2;
if (switchingRates.getDimension() != switchingRatesCount) {
throw new IllegalArgumentException(
"switching rates parameter must have "
+ switchingRatesCount
+ " dimensions, for "
+ hiddenStateCount
+ " hidden categories");
}
CovarionHKY model =
new CovarionHKY(dataType, kappaParam, hiddenClassRates, switchingRates, freqModel);
System.out.println(model);
return model;
}
public GMRFSkyrideLikelihood(
List<Tree> treeList,
Parameter popParameter,
Parameter groupParameter,
Parameter precParameter,
Parameter lambda,
Parameter beta,
MatrixParameter dMatrix,
boolean timeAwareSmoothing,
boolean rescaleByRootHeight) {
super(GMRFSkyrideLikelihoodParser.SKYLINE_LIKELIHOOD);
this.popSizeParameter = popParameter;
this.groupSizeParameter = groupParameter;
this.precisionParameter = precParameter;
this.lambdaParameter = lambda;
this.betaParameter = beta;
this.dMatrix = dMatrix;
this.timeAwareSmoothing = timeAwareSmoothing;
this.rescaleByRootHeight = rescaleByRootHeight;
addVariable(popSizeParameter);
addVariable(precisionParameter);
addVariable(lambdaParameter);
if (betaParameter != null) {
addVariable(betaParameter);
}
setTree(treeList);
int correctFieldLength = getCorrectFieldLength();
if (popSizeParameter.getDimension() <= 1) {
// popSize dimension hasn't been set yet, set it here:
popSizeParameter.setDimension(correctFieldLength);
}
fieldLength = popSizeParameter.getDimension();
if (correctFieldLength != fieldLength) {
throw new IllegalArgumentException(
"Population size parameter should have length " + correctFieldLength);
}
// Field length must be set by this point
wrapSetupIntervals();
coalescentIntervals = new double[fieldLength];
storedCoalescentIntervals = new double[fieldLength];
sufficientStatistics = new double[fieldLength];
storedSufficientStatistics = new double[fieldLength];
setupGMRFWeights();
addStatistic(new DeltaStatistic());
initializationReport();
/* Force all entries in groupSizeParameter = 1 for compatibility with Tracer */
if (groupSizeParameter != null) {
for (int i = 0; i < groupSizeParameter.getDimension(); i++)
groupSizeParameter.setParameterValue(i, 1.0);
}
}
public Object parseXMLObject(XMLObject xo) throws XMLParseException {
XMLObject cxo = xo.getChild(MVN_MEAN);
Parameter mean = (Parameter) cxo.getChild(Parameter.class);
cxo = xo.getChild(MVN_PRECISION);
MatrixParameter precision = (MatrixParameter) cxo.getChild(MatrixParameter.class);
if (mean.getDimension() != precision.getRowDimension()
|| mean.getDimension() != precision.getColumnDimension())
throw new XMLParseException(
"Mean and precision have wrong dimensions in " + xo.getName() + " element");
Transform[] transforms = parseListOfTransforms(xo, mean.getDimension());
MultivariateDistributionLikelihood likelihood =
new MultivariateDistributionLikelihood(
new MultivariateNormalDistribution(
mean.getParameterValues(), precision.getParameterAsMatrix()),
transforms);
cxo = xo.getChild(DATA);
if (cxo != null) {
for (int j = 0; j < cxo.getChildCount(); j++) {
if (cxo.getChild(j) instanceof Parameter) {
Parameter data = (Parameter) cxo.getChild(j);
if (data instanceof MatrixParameter) {
MatrixParameter matrix = (MatrixParameter) data;
if (matrix.getParameter(0).getDimension() != mean.getDimension())
throw new XMLParseException(
"dim("
+ data.getStatisticName()
+ ") = "
+ matrix.getParameter(0).getDimension()
+ " is not equal to dim("
+ mean.getStatisticName()
+ ") = "
+ mean.getDimension()
+ " in "
+ xo.getName()
+ "element");
for (int i = 0; i < matrix.getParameterCount(); i++) {
likelihood.addData(matrix.getParameter(i));
}
} else {
if (data.getDimension() != mean.getDimension())
throw new XMLParseException(
"dim("
+ data.getStatisticName()
+ ") = "
+ data.getDimension()
+ " is not equal to dim("
+ mean.getStatisticName()
+ ") = "
+ mean.getDimension()
+ " in "
+ xo.getName()
+ "element");
likelihood.addData(data);
}
} else {
throw new XMLParseException("illegal element in " + xo.getName() + " element");
}
}
}
return likelihood;
}
public Object parseXMLObject(XMLObject xo) throws XMLParseException {
XMLObject cxo = xo.getChild(DistributionLikelihoodParser.DISTRIBUTION);
ParametricMultivariateDistributionModel distribution =
(ParametricMultivariateDistributionModel)
cxo.getChild(ParametricMultivariateDistributionModel.class);
// Parse transforms here
int maxDim = distribution.getMean().length;
Transform[] transforms = parseListOfTransforms(xo, maxDim);
MultivariateDistributionLikelihood likelihood =
new MultivariateDistributionLikelihood(xo.getId(), distribution, transforms);
boolean dataAsMatrix = xo.getAttribute(DATA_AS_MATRIX, false);
cxo = xo.getChild(DATA);
if (cxo != null) {
for (int j = 0; j < cxo.getChildCount(); j++) {
if (cxo.getChild(j) instanceof Parameter) {
Parameter data = (Parameter) cxo.getChild(j);
if (data instanceof MatrixParameter) {
MatrixParameter matrix = (MatrixParameter) data;
if (dataAsMatrix) {
likelihood.addData(matrix);
} else {
if (matrix.getParameter(0).getDimension() != distribution.getMean().length)
throw new XMLParseException(
"dim("
+ data.getStatisticName()
+ ") = "
+ matrix.getParameter(0).getDimension()
+ " is not equal to dim("
+ distribution.getType()
+ ") = "
+ distribution.getMean().length
+ " in "
+ xo.getName()
+ "element");
for (int i = 0; i < matrix.getParameterCount(); i++) {
likelihood.addData(matrix.getParameter(i));
}
}
} else {
if (data.getDimension() != distribution.getMean().length)
throw new XMLParseException(
"dim("
+ data.getStatisticName()
+ ") = "
+ data.getDimension()
+ " is not equal to dim("
+ distribution.getType()
+ ") = "
+ distribution.getMean().length
+ " in "
+ xo.getName()
+ "element");
likelihood.addData(data);
}
} else {
throw new XMLParseException("illegal element in " + xo.getName() + " element");
}
}
}
return likelihood;
}
/** change the parameter and return the hastings ratio. */
public final double doOperation() {
int index = MathUtils.nextInt(links.getDimension());
int oldGroup = (int) assignments.getParameterValue(index);
/*
* Set index customer link to index and all connected to it to a new assignment (min value empty)
*/
int minEmp = minEmpty(modelLikelihood.getLogLikelihoodsVector());
links.setParameterValue(index, index);
int[] visited = connected(index, links);
int ii = 0;
while (visited[ii] != 0) {
assignments.setParameterValue(visited[ii] - 1, minEmp);
ii++;
}
/*
* Adjust likvector for group separated
*/
modelLikelihood.setLogLikelihoodsVector(oldGroup, getLogLikGroup(oldGroup));
modelLikelihood.setLogLikelihoodsVector(minEmp, getLogLikGroup(minEmp));
int maxFull = maxFull(modelLikelihood.getLogLikelihoodsVector());
double[] liks = modelLikelihood.getLogLikelihoodsVector();
/*
* computing likelihoods of joint groups
*/
double[] crossedLiks = new double[maxFull + 1];
for (int ll = 0; ll < maxFull + 1; ll++) {
if (ll != minEmp) {
crossedLiks[ll] = getLogLik2Group(ll, minEmp);
}
}
/*
* Add logPrior
*/
double[] logP = new double[links.getDimension()];
for (int jj = 0; jj < links.getDimension(); jj++) {
logP[jj] += depMatrix[index][jj];
int n = (int) assignments.getParameterValue(jj);
if (n != minEmp) {
logP[jj] += crossedLiks[n] - liks[n] - liks[minEmp];
}
}
logP[index] = Math.log(chiParameter.getParameterValue(0));
/*
* possibilidade de mandar p zero as probs muito pequenas
*/
/*
* Gibbs sampling
*/
this.rescale(logP); // Improve numerical stability
this.exp(logP); // Transform back to probability-scale
int k = MathUtils.randomChoicePDF(logP);
links.setParameterValue(index, k);
int newGroup = (int) assignments.getParameterValue(k);
ii = 0;
while (visited[ii] != 0) {
assignments.setParameterValue(visited[ii] - 1, newGroup);
ii++;
}
/*
* updating conditional likelihood vector
*/
modelLikelihood.setLogLikelihoodsVector(newGroup, getLogLikGroup(newGroup));
if (newGroup != minEmp) {
modelLikelihood.setLogLikelihoodsVector(minEmp, 0);
}
sampleMeans(maxFull);
return 0.0;
}
public void sampleMeans(int maxFull) {
double[][] means = new double[maxFull + 2][2];
// sample mean vector for each cluster
for (int i = 0; i < maxFull + 1; i++) {
// Find all elements in cluster
int ngroup = 0;
for (int ii = 0; ii < assignments.getDimension(); ii++) {
if ((int) assignments.getParameterValue(ii) == i) {
ngroup++;
}
}
if (ngroup != 0) {
double[][] group = new double[ngroup][2];
double[] groupMean = new double[2];
int count = 0;
for (int ii = 0; ii < assignments.getDimension(); ii++) {
if ((int) assignments.getParameterValue(ii) == i) {
group[count][0] = modelLikelihood.getData()[ii][0];
group[count][1] = modelLikelihood.getData()[ii][1];
groupMean[0] += group[count][0];
groupMean[1] += group[count][1];
count += 1;
}
}
groupMean[0] /= ngroup;
groupMean[1] /= ngroup;
double kn = k0 + ngroup;
double vn = v0 + ngroup;
double[][] sumdif = new double[2][2];
for (int jj = 0; jj < ngroup; jj++) {
sumdif[0][0] += (group[jj][0] - groupMean[0]) * (group[jj][0] - groupMean[0]);
sumdif[0][1] += (group[jj][0] - groupMean[0]) * (group[jj][1] - groupMean[1]);
sumdif[1][0] += (group[jj][0] - groupMean[0]) * (group[jj][1] - groupMean[1]);
sumdif[1][1] += (group[jj][1] - groupMean[1]) * (group[jj][1] - groupMean[1]);
}
double[][] TnInv = new double[2][2];
TnInv[0][0] =
T0Inv[0][0]
+ ngroup * (k0 / kn) * (groupMean[0] - m[0]) * (groupMean[0] - m[0])
+ sumdif[0][0];
TnInv[0][1] =
T0Inv[0][1]
+ ngroup * (k0 / kn) * (groupMean[1] - m[1]) * (groupMean[0] - m[0])
+ sumdif[0][1];
TnInv[1][0] =
T0Inv[1][0]
+ ngroup * (k0 / kn) * (groupMean[0] - m[0]) * (groupMean[1] - m[1])
+ sumdif[1][0];
TnInv[1][1] =
T0Inv[1][1]
+ ngroup * (k0 / kn) * (groupMean[1] - m[1]) * (groupMean[1] - m[1])
+ sumdif[1][1];
Matrix Tn = new SymmetricMatrix(TnInv).inverse();
double[] posteriorMean = new double[2];
// compute posterior mean
posteriorMean[0] = (k0 * m[0] + ngroup * groupMean[0]) / (k0 + ngroup);
posteriorMean[1] = (k0 * m[1] + ngroup * groupMean[1]) / (k0 + ngroup);
// compute posterior Precision
double[][] posteriorPrecision =
new WishartDistribution(vn, Tn.toComponents()).nextWishart();
posteriorPrecision[0][0] *= kn;
posteriorPrecision[1][0] *= kn;
posteriorPrecision[0][1] *= kn;
posteriorPrecision[1][1] *= kn;
double[] sample =
new MultivariateNormalDistribution(posteriorMean, posteriorPrecision)
.nextMultivariateNormal();
means[i][0] = sample[0];
means[i][1] = sample[1];
}
}
// Fill in cluster means for each observation
for (int j = 0; j < assignments.getDimension(); j++) {
double[] group = new double[2];
group[0] = means[(int) assignments.getParameterValue(j)][0];
group[1] = means[(int) assignments.getParameterValue(j)][1];
modelLikelihood.setMeans(j, group);
}
}
public double getLogLik2Group(int group1, int group2) {
double L = 0.0;
int ngroup1 = 0;
for (int i = 0; i < assignments.getDimension(); i++) {
if ((int) assignments.getParameterValue(i) == group1) {
ngroup1++;
}
}
int ngroup2 = 0;
for (int i = 0; i < assignments.getDimension(); i++) {
if ((int) assignments.getParameterValue(i) == group2) {
ngroup2++;
}
}
int ngroup = (ngroup1 + ngroup2);
if (ngroup != 0) {
double[][] group = new double[ngroup][2];
double mean[] = new double[2];
int count = 0;
for (int i = 0; i < assignments.getDimension(); i++) {
if ((int) assignments.getParameterValue(i) == group1) {
group[count][0] = modelLikelihood.getData()[i][0];
group[count][1] = modelLikelihood.getData()[i][1];
mean[0] += group[count][0];
mean[1] += group[count][1];
count += 1;
}
}
for (int i = 0; i < assignments.getDimension(); i++) {
if ((int) assignments.getParameterValue(i) == group2) {
group[count][0] = modelLikelihood.getData()[i][0];
group[count][1] = modelLikelihood.getData()[i][1];
mean[0] += group[count][0];
mean[1] += group[count][1];
count += 1;
}
}
mean[0] /= ngroup;
mean[1] /= ngroup;
double kn = k0 + ngroup;
double vn = v0 + ngroup;
double[][] sumdif = new double[2][2];
for (int i = 0; i < ngroup; i++) {
sumdif[0][0] += (group[i][0] - mean[0]) * (group[i][0] - mean[0]);
sumdif[0][1] += (group[i][0] - mean[0]) * (group[i][1] - mean[1]);
sumdif[1][0] += (group[i][0] - mean[0]) * (group[i][1] - mean[1]);
sumdif[1][1] += (group[i][1] - mean[1]) * (group[i][1] - mean[1]);
}
double[][] TnInv = new double[2][2];
TnInv[0][0] =
T0Inv[0][0] + ngroup * (k0 / kn) * (mean[0] - m[0]) * (mean[0] - m[0]) + sumdif[0][0];
TnInv[0][1] =
T0Inv[0][1] + ngroup * (k0 / kn) * (mean[1] - m[1]) * (mean[0] - m[0]) + sumdif[0][1];
TnInv[1][0] =
T0Inv[1][0] + ngroup * (k0 / kn) * (mean[0] - m[0]) * (mean[1] - m[1]) + sumdif[1][0];
TnInv[1][1] =
T0Inv[1][1] + ngroup * (k0 / kn) * (mean[1] - m[1]) * (mean[1] - m[1]) + sumdif[1][1];
double logDetTn = -Math.log(TnInv[0][0] * TnInv[1][1] - TnInv[0][1] * TnInv[1][0]);
L += -(ngroup) * Math.log(Math.PI);
L += Math.log(k0) - Math.log(kn);
L += (vn / 2) * logDetTn - (v0 / 2) * logDetT0;
L += GammaFunction.lnGamma(vn / 2) + GammaFunction.lnGamma((vn / 2) - 0.5);
L += -GammaFunction.lnGamma(v0 / 2) - GammaFunction.lnGamma((v0 / 2) - 0.5);
}
return L;
}
public Object parseXMLObject(XMLObject xo) throws XMLParseException {
Parameter ratesParameter = null;
FrequencyModel freqModel = null;
if (xo.hasChildNamed(FREQUENCIES)) {
XMLObject cxo = xo.getChild(FREQUENCIES);
freqModel = (FrequencyModel) cxo.getChild(FrequencyModel.class);
}
DataType dataType = DataTypeUtils.getDataType(xo);
if (dataType == null) dataType = (DataType) xo.getChild(DataType.class);
// if (xo.hasAttribute(DataType.DATA_TYPE)) {
// String dataTypeStr = xo.getStringAttribute(DataType.DATA_TYPE);
// if (dataTypeStr.equals(Nucleotides.DESCRIPTION)) {
// dataType = Nucleotides.INSTANCE;
// } else if (dataTypeStr.equals(AminoAcids.DESCRIPTION)) {
// dataType = AminoAcids.INSTANCE;
// } else if (dataTypeStr.equals(Codons.DESCRIPTION)) {
// dataType = Codons.UNIVERSAL;
// } else if (dataTypeStr.equals(TwoStates.DESCRIPTION)) {
// dataType = TwoStates.INSTANCE;
// }
// }
if (dataType == null) dataType = freqModel.getDataType();
if (dataType != freqModel.getDataType()) {
throw new XMLParseException(
"Data type of "
+ getParserName()
+ " element does not match that of its frequencyModel.");
}
XMLObject cxo = xo.getChild(RATES);
ratesParameter = (Parameter) cxo.getChild(Parameter.class);
int states = dataType.getStateCount();
Logger.getLogger("dr.evomodel")
.info(" General Substitution Model (stateCount=" + states + ")");
boolean hasRelativeRates =
cxo.hasChildNamed(RELATIVE_TO)
|| (cxo.hasAttribute(RELATIVE_TO) && cxo.getIntegerAttribute(RELATIVE_TO) > 0);
int nonReversibleRateCount = ((dataType.getStateCount() - 1) * dataType.getStateCount());
int reversibleRateCount = (nonReversibleRateCount / 2);
boolean isNonReversible = ratesParameter.getDimension() == nonReversibleRateCount;
boolean hasIndicator = xo.hasChildNamed(INDICATOR);
if (!hasRelativeRates) {
Parameter indicatorParameter = null;
if (ratesParameter.getDimension() != reversibleRateCount
&& ratesParameter.getDimension() != nonReversibleRateCount) {
throw new XMLParseException(
"Rates parameter in "
+ getParserName()
+ " element should have "
+ (reversibleRateCount)
+ " dimensions for reversible model or "
+ nonReversibleRateCount
+ " dimensions for non-reversible. "
+ "However parameter dimension is "
+ ratesParameter.getDimension());
}
if (hasIndicator) { // this is using BSSVS
cxo = xo.getChild(INDICATOR);
indicatorParameter = (Parameter) cxo.getChild(Parameter.class);
if (indicatorParameter.getDimension() != ratesParameter.getDimension()) {
throw new XMLParseException(
"Rates and indicator parameters in "
+ getParserName()
+ " element must be the same dimension.");
}
boolean randomize =
xo.getAttribute(
dr.evomodelxml.substmodel.ComplexSubstitutionModelParser.RANDOMIZE, false);
if (randomize) {
BayesianStochasticSearchVariableSelection.Utils.randomize(
indicatorParameter, dataType.getStateCount(), !isNonReversible);
}
}
if (isNonReversible) {
// if (xo.hasChildNamed(ROOT_FREQ)) {
// cxo = xo.getChild(ROOT_FREQ);
// FrequencyModel rootFreq = (FrequencyModel)
// cxo.getChild(FrequencyModel.class);
//
// if (dataType != rootFreq.getDataType()) {
// throw new XMLParseException("Data type of " + getParserName() + "
// element does not match that of its rootFrequencyModel.");
// }
//
// Logger.getLogger("dr.evomodel").info(" Using BSSVS Complex
// Substitution Model");
// return new SVSComplexSubstitutionModel(getParserName(), dataType,
// freqModel, ratesParameter, indicatorParameter);
//
// } else {
// throw new XMLParseException("Non-reversible model missing " +
// ROOT_FREQ + " element");
// }
Logger.getLogger("dr.evomodel").info(" Using BSSVS Complex Substitution Model");
return new SVSComplexSubstitutionModel(
getParserName(), dataType, freqModel, ratesParameter, indicatorParameter);
} else {
Logger.getLogger("dr.evomodel").info(" Using BSSVS General Substitution Model");
return new SVSGeneralSubstitutionModel(
getParserName(), dataType, freqModel, ratesParameter, indicatorParameter);
}
} else {
// if we have relativeTo attribute then we use the old GeneralSubstitutionModel
if (ratesParameter.getDimension() != reversibleRateCount - 1) {
throw new XMLParseException(
"Rates parameter in "
+ getParserName()
+ " element should have "
+ (reversibleRateCount - 1)
+ " dimensions. However parameter dimension is "
+ ratesParameter.getDimension());
}
int relativeTo = 0;
if (hasRelativeRates) {
relativeTo = cxo.getIntegerAttribute(RELATIVE_TO) - 1;
}
if (relativeTo < 0 || relativeTo >= reversibleRateCount) {
throw new XMLParseException(RELATIVE_TO + " must be 1 or greater");
} else {
int t = relativeTo;
int s = states - 1;
int row = 0;
while (t >= s) {
t -= s;
s -= 1;
row += 1;
}
int col = t + row + 1;
Logger.getLogger("dr.evomodel")
.info(" Rates relative to " + dataType.getCode(row) + "<->" + dataType.getCode(col));
}
if (ratesParameter == null) {
if (reversibleRateCount == 1) {
// simplest model for binary traits...
} else {
throw new XMLParseException("No rates parameter found in " + getParserName());
}
}
return new GeneralSubstitutionModel(
getParserName(), dataType, freqModel, ratesParameter, relativeTo);
}
}
public Object parseXMLObject(XMLObject xo) throws XMLParseException {
double[] shape;
double[] scale;
if (xo.hasChildNamed(MVGAMMA_SHAPE)) {
XMLObject cxo = xo.getChild(MVGAMMA_SHAPE);
shape = ((Parameter) cxo.getChild(Parameter.class)).getParameterValues();
cxo = xo.getChild(MVGAMMA_SCALE);
scale = ((Parameter) cxo.getChild(Parameter.class)).getParameterValues();
if (shape.length != scale.length)
throw new XMLParseException(
"Shape and scale have wrong dimensions in " + xo.getName() + " element");
} else {
XMLObject cxo = xo.getChild(MVN_MEAN);
double[] mean = ((Parameter) cxo.getChild(Parameter.class)).getParameterValues();
cxo = xo.getChild(MVN_CV);
double[] cv = ((Parameter) cxo.getChild(Parameter.class)).getParameterValues();
if (mean.length != cv.length)
throw new XMLParseException(
"Mean and CV have wrong dimensions in " + xo.getName() + " element");
final int dim = mean.length;
shape = new double[dim];
scale = new double[dim];
for (int i = 0; i < dim; i++) {
double c2 = cv[i] * cv[i];
shape[i] = 1.0 / c2;
scale[i] = c2 * mean[i];
}
}
MultivariateDistributionLikelihood likelihood =
new MultivariateDistributionLikelihood(
new MultivariateGammaDistribution(shape, scale));
XMLObject cxo = xo.getChild(DATA);
for (int j = 0; j < cxo.getChildCount(); j++) {
if (cxo.getChild(j) instanceof Parameter) {
Parameter data = (Parameter) cxo.getChild(j);
likelihood.addData(data);
if (data.getDimension() != shape.length)
throw new XMLParseException(
"dim("
+ data.getStatisticName()
+ ") != "
+ shape.length
+ " in "
+ xo.getName()
+ "element");
} else {
throw new XMLParseException("illegal element in " + xo.getName() + " element");
}
}
return likelihood;
}
public void initializationReport() {
System.out.println("Creating a GMRF smoothed skyride model:");
System.out.println("\tPopulation sizes: " + popSizeParameter.getDimension());
System.out.println(
"\tIf you publish results using this model, please reference: Minin, Bloomquist and Suchard (2008) Molecular Biology and Evolution, 25, 1459-1471.");
}
public MixtureModelBranchRates(
TreeModel tree,
Parameter rateCategoryQuantilesParameter,
ParametricDistributionModel[] models,
Parameter distributionIndexParameter,
boolean useQuantilesForRates,
boolean normalize,
double normalizeBranchRateTo) {
super(MixtureModelBranchRatesParser.MIXTURE_MODEL_BRANCH_RATES);
this.useQuantilesForRates = useQuantilesForRates;
this.rateCategoryQuantiles =
new TreeParameterModel(tree, rateCategoryQuantilesParameter, false);
rates = new double[tree.getNodeCount()];
this.normalize = normalize;
this.treeModel = tree;
this.distributionModels = models;
this.normalizeBranchRateTo = normalizeBranchRateTo;
this.tree = new SimpleTree(tree);
this.distributionIndexParameter = distributionIndexParameter;
addVariable(this.distributionIndexParameter);
// Force the boundaries of rateCategoryParameter to match the category count
// d Parameter.DefaultBounds bound = new Parameter.DefaultBounds(categoryCount - 1, 0,
// rateCategoryParameter.getDimension());
// d rateCategoryParameter.addBounds(bound);
// rateCategoryQuantilesParameter.;
Parameter.DefaultBounds bound =
new Parameter.DefaultBounds(1.0, 0.0, rateCategoryQuantilesParameter.getDimension());
rateCategoryQuantilesParameter.addBounds(bound);
Parameter.DefaultBounds bound2 = new Parameter.DefaultBounds(models.length, 0.0, 1);
distributionIndexParameter.addBounds(bound2);
distributionIndexParameter.setParameterValue(0, 0);
// Parameter distributionIndexParameter;
for (ParametricDistributionModel distributionModel : distributionModels) {
addModel(distributionModel);
}
// AR - commented out: changes to the tree are handled by model changed events fired by
// rateCategories
// addModel(tree);
// d addModel(rateCategories);
addModel(rateCategoryQuantiles);
// addModel(treeModel); // Maybe
// AR - commented out: changes to rateCategoryParameter are handled by model changed events
// fired by rateCategories
// addVariable(rateCategoryParameter);
if (normalize) {
tree.addModelListener(
new ModelListener() {
public void modelChangedEvent(Model model, Object object, int index) {
computeFactor();
}
public void modelRestored(Model model) {
computeFactor();
}
});
}
setupRates();
}