/*
* TreePriorGenerator.java
*
* Copyright (c) 2002-2015 Alexei Drummond, Andrew Rambaut and Marc Suchard
*
* This file is part of BEAST.
* See the NOTICE file distributed with this work for additional
* information regarding copyright ownership and licensing.
*
* BEAST is free software; you can redistribute it and/or modify
* it under the terms of the GNU Lesser General Public License as
* published by the Free Software Foundation; either version 2
* of the License, or (at your option) any later version.
*
* BEAST 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 Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public
* License along with BEAST; if not, write to the
* Free Software Foundation, Inc., 51 Franklin St, Fifth Floor,
* Boston, MA 02110-1301 USA
*/
package dr.app.beauti.generator;
import dr.app.beauti.components.ComponentFactory;
import dr.app.beauti.options.*;
import dr.app.beauti.types.StartingTreeType;
import dr.app.beauti.types.TreePriorParameterizationType;
import dr.app.beauti.types.TreePriorType;
import dr.app.beauti.util.XMLWriter;
import dr.evolution.util.Taxa;
import dr.evolution.util.Units;
import dr.evomodel.tree.TreeModel;
import dr.evomodelxml.CSVExporterParser;
import dr.evomodelxml.coalescent.*;
import dr.evomodelxml.speciation.*;
import dr.evoxml.TaxaParser;
import dr.inference.distribution.ExponentialDistributionModel;
import dr.inference.distribution.ExponentialMarkovModel;
import dr.inference.model.ParameterParser;
import dr.inferencexml.distribution.DistributionModelParser;
import dr.inferencexml.distribution.ExponentialMarkovModelParser;
import dr.inferencexml.distribution.MixedDistributionLikelihoodParser;
import dr.inferencexml.model.SumStatisticParser;
import dr.util.Attribute;
import dr.xml.XMLParser;
/**
* @author Alexei Drummond
* @author Andrew Rambaut
*/
public class TreePriorGenerator extends Generator {
public TreePriorGenerator(BeautiOptions options, ComponentFactory[] components) {
super(options, components);
}
// void writeTreePrior(PartitionTreePrior prior, PartitionTreeModel model, XMLWriter writer) { // for species, partitionName.treeModel
// setModelPrefix(prior.getPrefix()); // only has prefix, if (options.getPartitionTreePriors().size() > 1)
//
// writePriorLikelihood(prior, model, writer);
// }
/**
* Write a tree prior (coalescent or speciational) model
*
* @param prior the partition tree prior
* @param writer the writer
*/
void writeTreePriorModel(PartitionTreePrior prior, XMLWriter writer) {
setModelPrefix(prior.getPrefix()); // only has prefix, if (options.getPartitionTreePriors().size() > 1)
String initialPopSize = null;
TreePriorType nodeHeightPrior = prior.getNodeHeightPrior();
Units.Type units = options.units;
TreePriorParameterizationType parameterization = prior.getParameterization();
switch (nodeHeightPrior) {
case CONSTANT:
writer.writeComment("A prior assumption that the population size has remained constant",
"throughout the time spanned by the genealogy.");
writer.writeOpenTag(
ConstantPopulationModelParser.CONSTANT_POPULATION_MODEL,
new Attribute[]{
new Attribute.Default<String>(XMLParser.ID, modelPrefix + "constant"),
new Attribute.Default<String>("units", Units.Utils.getDefaultUnitName(options.units))
}
);
writer.writeOpenTag(ConstantPopulationModelParser.POPULATION_SIZE);
writeParameter("constant.popSize", prior, writer);
writer.writeCloseTag(ConstantPopulationModelParser.POPULATION_SIZE);
writer.writeCloseTag(ConstantPopulationModelParser.CONSTANT_POPULATION_MODEL);
break;
case EXPONENTIAL:
// generate an exponential prior tree
writer.writeComment("A prior assumption that the population size has grown exponentially",
"throughout the time spanned by the genealogy.");
writer.writeOpenTag(
ExponentialGrowthModelParser.EXPONENTIAL_GROWTH_MODEL,
new Attribute[]{
new Attribute.Default<String>(XMLParser.ID, modelPrefix + "exponential"),
new Attribute.Default<String>("units", Units.Utils.getDefaultUnitName(options.units))
}
);
// write pop size socket
writer.writeOpenTag(ExponentialGrowthModelParser.POPULATION_SIZE);
writeParameter("exponential.popSize", prior, writer);
writer.writeCloseTag(ExponentialGrowthModelParser.POPULATION_SIZE);
if (parameterization == TreePriorParameterizationType.GROWTH_RATE) {
// write growth rate socket
writer.writeOpenTag(ExponentialGrowthModelParser.GROWTH_RATE);
writeParameter("exponential.growthRate", prior, writer);
writer.writeCloseTag(ExponentialGrowthModelParser.GROWTH_RATE);
} else {
// write doubling time socket
writer.writeOpenTag(ExponentialGrowthModelParser.DOUBLING_TIME);
writeParameter("exponential.doublingTime", prior, writer);
writer.writeCloseTag(ExponentialGrowthModelParser.DOUBLING_TIME);
}
writer.writeCloseTag(ExponentialGrowthModelParser.EXPONENTIAL_GROWTH_MODEL);
break;
case LOGISTIC:
// generate an exponential prior tree
writer.writeComment("A prior assumption that the population size has grown logistically",
"throughout the time spanned by the genealogy.");
writer.writeOpenTag(
LogisticGrowthModelParser.LOGISTIC_GROWTH_MODEL,
new Attribute[]{
new Attribute.Default<String>(XMLParser.ID, modelPrefix + "logistic"),
new Attribute.Default<String>("units", Units.Utils.getDefaultUnitName(options.units))
}
);
// write pop size socket
writer.writeOpenTag(LogisticGrowthModelParser.POPULATION_SIZE);
writeParameter("logistic.popSize", prior, writer);
writer.writeCloseTag(LogisticGrowthModelParser.POPULATION_SIZE);
if (parameterization == TreePriorParameterizationType.GROWTH_RATE) {
// write growth rate socket
writer.writeOpenTag(LogisticGrowthModelParser.GROWTH_RATE);
writeParameter("logistic.growthRate", prior, writer);
writer.writeCloseTag(LogisticGrowthModelParser.GROWTH_RATE);
} else {
// write doubling time socket
writer.writeOpenTag(LogisticGrowthModelParser.DOUBLING_TIME);
writeParameter("logistic.doublingTime", prior, writer);
writer.writeCloseTag(LogisticGrowthModelParser.DOUBLING_TIME);
}
// write logistic t50 socket
writer.writeOpenTag(LogisticGrowthModelParser.TIME_50);
// if (options.clockModelOptions.getRateOptionClockModel() == FixRateType.FIX_MEAN
// || options.clockModelOptions.getRateOptionClockModel() == FixRateType.RELATIVE_TO) {
// writer.writeComment("No calibration");
// writer.writeComment("logistic.t50 initial always has to < treeRootHeight initial");
// dr.app.beauti.options.Parameter priorPara = prior.getParameter("logistic.t50");
//
// double initRootHeight;
// if (options.isShareSameTreePrior()) {
// initRootHeight = priorPara.initial;
// for (PartitionTreeModel tree : options.getPartitionTreeModels()) {
// double tmpRootHeight = tree.getParameter("treeModel.rootHeight").initial;
// if (initRootHeight > tmpRootHeight) { // take min
// initRootHeight = tmpRootHeight;
// }
// }
// } else {
// initRootHeight = prior.getTreeModel().getParameter("treeModel.rootHeight").initial;
// }
// // logistic.t50 initial always has to < treeRootHeight initial
// if (priorPara.initial >= initRootHeight) {
// priorPara.initial = initRootHeight / 2; // tree prior.initial has to < treeRootHeight.initial
// }
// } else {
// writer.writeComment("Has calibration");
//
// throw new IllegalArgumentException("This function is not available in this release !");
// }
writeParameter("logistic.t50", prior, writer);
writer.writeCloseTag(LogisticGrowthModelParser.TIME_50);
writer.writeCloseTag(LogisticGrowthModelParser.LOGISTIC_GROWTH_MODEL);
initialPopSize = "logistic.popSize";
break;
case EXPANSION:
// generate an exponential prior tree
writer.writeComment("A prior assumption that the population size has grown exponentially",
"from some ancestral population size in the past.");
writer.writeOpenTag(
ExpansionModelParser.EXPANSION_MODEL,
new Attribute[]{
new Attribute.Default<String>(XMLParser.ID, modelPrefix + "expansion"),
new Attribute.Default<String>("units", Units.Utils.getDefaultUnitName(options.units))
}
);
// write pop size socket
writeParameter(ExpansionModelParser.POPULATION_SIZE, "expansion.popSize", prior, writer);
if (parameterization == TreePriorParameterizationType.GROWTH_RATE) {
// write growth rate socket
writeParameter(ExpansionModelParser.GROWTH_RATE, "expansion.growthRate", prior, writer);
} else {
// write doubling time socket
writeParameter(ExpansionModelParser.DOUBLING_TIME, "expansion.doublingTime", prior, writer);
}
// write ancestral proportion socket
writeParameter(ExpansionModelParser.ANCESTRAL_POPULATION_PROPORTION, "expansion.ancestralProportion", prior, writer);
writer.writeCloseTag(ExpansionModelParser.EXPANSION_MODEL);
initialPopSize = "expansion.popSize";
break;
case YULE:
case YULE_CALIBRATION:
if (nodeHeightPrior == TreePriorType.YULE_CALIBRATION) {
writer.writeComment("Calibrated Yule: Heled J, Drummond AJ (2011), Syst Biol, doi: " +
"10.1093/sysbio/syr087");
} else {
writer.writeComment("A prior on the distribution node heights defined given",
"a Yule speciation process (a pure birth process).");
}
writer.writeOpenTag(
YuleModelParser.YULE_MODEL,
new Attribute[]{
new Attribute.Default<String>(XMLParser.ID, modelPrefix + YuleModelParser.YULE),
new Attribute.Default<String>("units", Units.Utils.getDefaultUnitName(units))
}
);
writeParameter(YuleModelParser.BIRTH_RATE, "yule.birthRate", prior, writer);
writer.writeCloseTag(YuleModelParser.YULE_MODEL);
break;
case BIRTH_DEATH:
case BIRTH_DEATH_INCOMPLETE_SAMPLING:
writer.writeComment("A prior on the distribution node heights defined given");
writer.writeComment(nodeHeightPrior == TreePriorType.BIRTH_DEATH_INCOMPLETE_SAMPLING ?
BirthDeathModelParser.getCitationRHO() : BirthDeathModelParser.getCitation());
writer.writeOpenTag(
BirthDeathModelParser.BIRTH_DEATH_MODEL,
new Attribute[]{
new Attribute.Default<String>(XMLParser.ID, modelPrefix + BirthDeathModelParser.BIRTH_DEATH),
new Attribute.Default<String>("units", Units.Utils.getDefaultUnitName(units))
}
);
writeParameter(BirthDeathModelParser.BIRTHDIFF_RATE, BirthDeathModelParser.MEAN_GROWTH_RATE_PARAM_NAME, prior, writer);
writeParameter(BirthDeathModelParser.RELATIVE_DEATH_RATE, BirthDeathModelParser.RELATIVE_DEATH_RATE_PARAM_NAME, prior, writer);
if (nodeHeightPrior == TreePriorType.BIRTH_DEATH_INCOMPLETE_SAMPLING) {
writeParameter(BirthDeathModelParser.SAMPLE_PROB,
BirthDeathModelParser.BIRTH_DEATH + "." + BirthDeathModelParser.SAMPLE_PROB, prior, writer);
}
writer.writeCloseTag(BirthDeathModelParser.BIRTH_DEATH_MODEL);
break;
case BIRTH_DEATH_SERIAL_SAMPLING:
writer.writeComment(BirthDeathSerialSamplingModelParser.getCitationPsiOrg());
writer.writeOpenTag(
BirthDeathSerialSamplingModelParser.BIRTH_DEATH_SERIAL_MODEL,
new Attribute[]{
new Attribute.Default<String>(XMLParser.ID, modelPrefix + BirthDeathSerialSamplingModelParser.BDSS),
new Attribute.Default<String>("units", Units.Utils.getDefaultUnitName(units)),
new Attribute.Default<Boolean>(BirthDeathSerialSamplingModelParser.HAS_FINAL_SAMPLE, false)
}
);
writeParameter(BirthDeathSerialSamplingModelParser.LAMBDA,
BirthDeathSerialSamplingModelParser.BDSS + "." + BirthDeathSerialSamplingModelParser.LAMBDA, prior, writer);
writeParameter(BirthDeathSerialSamplingModelParser.RELATIVE_MU,
BirthDeathSerialSamplingModelParser.BDSS + "." + BirthDeathSerialSamplingModelParser.RELATIVE_MU, prior, writer);
// writeParameter(BirthDeathSerialSamplingModelParser.SAMPLE_PROBABILITY,
// BirthDeathSerialSamplingModelParser.BDSS + "." + BirthDeathSerialSamplingModelParser.SAMPLE_PROBABILITY, prior, writer);
writeParameter(BirthDeathSerialSamplingModelParser.PSI,
BirthDeathSerialSamplingModelParser.BDSS + "." + BirthDeathSerialSamplingModelParser.PSI, prior, writer);
writeParameter(BirthDeathSerialSamplingModelParser.ORIGIN,
BirthDeathSerialSamplingModelParser.BDSS + "." + BirthDeathSerialSamplingModelParser.ORIGIN, prior, writer);
writer.writeCloseTag(BirthDeathSerialSamplingModelParser.BIRTH_DEATH_SERIAL_MODEL);
break;
case BIRTH_DEATH_BASIC_REPRODUCTIVE_NUMBER:
writer.writeComment(BirthDeathSerialSamplingModelParser.getCitationRT());
writer.writeOpenTag(
BirthDeathEpidemiologyModelParser.BIRTH_DEATH_EPIDEMIOLOGY,
new Attribute[]{
new Attribute.Default<String>(XMLParser.ID, modelPrefix + BirthDeathEpidemiologyModelParser.BIRTH_DEATH_EPIDEMIOLOGY),
new Attribute.Default<String>("units", Units.Utils.getDefaultUnitName(units))
}
);
writeParameter(BirthDeathEpidemiologyModelParser.R0,
BirthDeathEpidemiologyModelParser.R0, prior, writer);
writeParameter(BirthDeathEpidemiologyModelParser.RECOVERY_RATE,
BirthDeathEpidemiologyModelParser.RECOVERY_RATE, prior, writer);
writeParameter(BirthDeathEpidemiologyModelParser.SAMPLING_PROBABILITY,
BirthDeathEpidemiologyModelParser.SAMPLING_PROBABILITY, prior, writer);
writeParameter(BirthDeathEpidemiologyModelParser.ORIGIN,
BirthDeathEpidemiologyModelParser.ORIGIN, prior, writer);
writer.writeCloseTag(BirthDeathEpidemiologyModelParser.BIRTH_DEATH_EPIDEMIOLOGY);
break;
case SPECIES_BIRTH_DEATH:
case SPECIES_YULE:
case SPECIES_YULE_CALIBRATION:
writer.writeComment("A prior assumption that the population size has remained constant");
writer.writeComment("throughout the time spanned by the genealogy.");
if (nodeHeightPrior == TreePriorType.SPECIES_YULE_CALIBRATION)
writer.writeComment("Calibrated Yule: Heled J, Drummond AJ (2011), Syst Biol, doi: " +
"10.1093/sysbio/syr087");
writer.writeOpenTag(
ConstantPopulationModelParser.CONSTANT_POPULATION_MODEL,
new Attribute[]{
new Attribute.Default<String>(XMLParser.ID, modelPrefix + "constant"),
new Attribute.Default<String>("units", Units.Utils.getDefaultUnitName(options.units))
}
);
// initial value for pop mean is the same as what used to be the value for the population size
Parameter para = options.starBEASTOptions.getParameter(TraitData.TRAIT_SPECIES + "." + options.starBEASTOptions.POP_MEAN);
prior.getParameter("constant.popSize").setInitial(para.getInitial());
writer.writeOpenTag(ConstantPopulationModelParser.POPULATION_SIZE);
writeParameter("constant.popSize", prior, writer);
writer.writeCloseTag(ConstantPopulationModelParser.POPULATION_SIZE);
writer.writeCloseTag(ConstantPopulationModelParser.CONSTANT_POPULATION_MODEL);
break;
}
if ((!options.useStarBEAST) && nodeHeightPrior != TreePriorType.CONSTANT && nodeHeightPrior != TreePriorType.EXPONENTIAL) {
// If the node height prior is not one of these two then we need to simulate a
// random starting tree under a constant size coalescent.
writer.writeComment("This is a simple constant population size coalescent model",
"that is used to generate an initial tree for the chain.");
writer.writeOpenTag(
ConstantPopulationModelParser.CONSTANT_POPULATION_MODEL,
new Attribute[]{
new Attribute.Default<String>(XMLParser.ID, modelPrefix + "initialDemo"),
new Attribute.Default<String>("units", Units.Utils.getDefaultUnitName(units))
}
);
writer.writeOpenTag(ConstantPopulationModelParser.POPULATION_SIZE);
if (initialPopSize != null) {
writer.writeIDref(ParameterParser.PARAMETER, modelPrefix + initialPopSize);
} else {
writeParameter(modelPrefix + "initialDemo.popSize", 1, 100.0, Double.NaN, Double.NaN, writer);
}
writer.writeCloseTag(ConstantPopulationModelParser.POPULATION_SIZE);
writer.writeCloseTag(ConstantPopulationModelParser.CONSTANT_POPULATION_MODEL);
}
// if (nodeHeightPrior == TreePriorType.BIRTH_DEATH_BASIC_REPRODUCTIVE_NUMBER) {
// writer.writeComment("R0 = b/(b*d+s*r)");
// writer.writeOpenTag(RPNcalculatorStatisticParser.RPN_STATISTIC,
// new Attribute[]{
// new Attribute.Default<String>(XMLParser.ID, modelPrefix + "R0")
// });
//
// writer.writeOpenTag(RPNcalculatorStatisticParser.VARIABLE,
// new Attribute[]{
// new Attribute.Default<String>(Statistic.NAME, modelPrefix + "b")
// });
// writeParameterRef(modelPrefix + BirthDeathSerialSamplingModelParser.BDSS + "." + BirthDeathSerialSamplingModelParser.LAMBDA, writer);
// writer.writeCloseTag(RPNcalculatorStatisticParser.VARIABLE);
//
// writer.writeOpenTag(RPNcalculatorStatisticParser.VARIABLE,
// new Attribute[]{
// new Attribute.Default<String>(Statistic.NAME, modelPrefix + "d")
// });
// writeParameterRef(modelPrefix + BirthDeathSerialSamplingModelParser.BDSS + "." + BirthDeathSerialSamplingModelParser.RELATIVE_MU, writer);
// writer.writeCloseTag(RPNcalculatorStatisticParser.VARIABLE);
//
// writer.writeOpenTag(RPNcalculatorStatisticParser.VARIABLE,
// new Attribute[]{
// new Attribute.Default<String>(Statistic.NAME, modelPrefix + "s")
// });
// writeParameterRef(modelPrefix + BirthDeathSerialSamplingModelParser.BDSS + "." + BirthDeathSerialSamplingModelParser.PSI, writer);
// writer.writeCloseTag(RPNcalculatorStatisticParser.VARIABLE);
//
// writer.writeOpenTag(RPNcalculatorStatisticParser.VARIABLE,
// new Attribute[]{
// new Attribute.Default<String>(Statistic.NAME, modelPrefix + "r")
// });
// writeParameterRef(modelPrefix + BirthDeathSerialSamplingModelParser.BDSS + "." + BirthDeathSerialSamplingModelParser.R, writer);
// writer.writeCloseTag(RPNcalculatorStatisticParser.VARIABLE);
//
// writer.writeOpenTag(RPNcalculatorStatisticParser.EXPRESSION,
// new Attribute[]{
// new Attribute.Default<String>(Statistic.NAME, modelPrefix + "R0")
// });
// writer.writeText(modelPrefix + "b " + modelPrefix + "b " + modelPrefix + "d " + "* " + modelPrefix + "s " + modelPrefix + "r " + "* + /");
// writer.writeCloseTag(RPNcalculatorStatisticParser.EXPRESSION);
//
// writer.writeCloseTag(RPNcalculatorStatisticParser.RPN_STATISTIC);
// }
}
/**
* Write the prior on node heights (coalescent or speciational models)
*
* @param model PartitionTreeModel
* @param writer the writer
*/
void writePriorLikelihood(PartitionTreeModel model, XMLWriter writer) {
//tree model prefix
setModelPrefix(model.getPrefix()); // only has prefix, if (options.getPartitionTreePriors().size() > 1)
// String priorPrefix = prior.getPrefix();
PartitionTreePrior prior = model.getPartitionTreePrior();
TreePriorType treePrior = prior.getNodeHeightPrior();
switch (treePrior) {
case YULE:
case BIRTH_DEATH:
case BIRTH_DEATH_INCOMPLETE_SAMPLING:
case BIRTH_DEATH_SERIAL_SAMPLING:
case BIRTH_DEATH_BASIC_REPRODUCTIVE_NUMBER:
case YULE_CALIBRATION:
// generate a speciational process
writer.writeComment("Generate a speciation likelihood for Yule or Birth Death");
writer.writeOpenTag(
SpeciationLikelihoodParser.SPECIATION_LIKELIHOOD,
new Attribute[]{
new Attribute.Default<String>(XMLParser.ID, modelPrefix + "speciation")
}
);
// write pop size socket
writer.writeOpenTag(SpeciationLikelihoodParser.MODEL);
writeNodeHeightPriorModelRef(prior, writer);
writer.writeCloseTag(SpeciationLikelihoodParser.MODEL);
writer.writeOpenTag(SpeciationLikelihoodParser.TREE);
writer.writeIDref(TreeModel.TREE_MODEL, modelPrefix + TreeModel.TREE_MODEL);
writer.writeCloseTag(SpeciationLikelihoodParser.TREE);
if (treePrior == TreePriorType.YULE_CALIBRATION) {
if (options.treeModelOptions.isNodeCalibrated(model) == 0) {
writer.writeOpenTag(SpeciationLikelihoodParser.CALIBRATION,
new Attribute[]{
new Attribute.Default<String>(SpeciationLikelihoodParser.CORRECTION, prior.getCalibCorrectionType().toString())
});
writer.writeOpenTag(SpeciationLikelihoodParser.POINT);
String taxaId;
if (options.hasIdenticalTaxa()) {
taxaId = TaxaParser.TAXA;
} else {
taxaId = options.getDataPartitions(model).get(0).getPrefix() + TaxaParser.TAXA;
}
writer.writeIDref(TaxaParser.TAXA, taxaId);
writeDistribution(model.getParameter("treeModel.rootHeight"), true, writer);
writer.writeCloseTag(SpeciationLikelihoodParser.POINT);
writer.writeCloseTag(SpeciationLikelihoodParser.CALIBRATION);
} else if (options.treeModelOptions.isNodeCalibrated(model) == 1) {
// should be only 1 calibrated internal node with monophyletic for each tree at moment
Taxa t = (Taxa) options.getKeysFromValue(options.taxonSetsTreeModel, model).get(0);
Parameter nodeCalib = options.getStatistic(t);
writer.writeOpenTag(SpeciationLikelihoodParser.CALIBRATION,
new Attribute[]{
new Attribute.Default<String>(SpeciationLikelihoodParser.CORRECTION, prior.getCalibCorrectionType().toString())
});
writer.writeOpenTag(SpeciationLikelihoodParser.POINT);
writer.writeIDref(TaxaParser.TAXA, t.getId());
writeDistribution(nodeCalib, true, writer);
writer.writeCloseTag(SpeciationLikelihoodParser.POINT);
writer.writeCloseTag(SpeciationLikelihoodParser.CALIBRATION);
if (!options.treeModelOptions.isNodeCalibrated(nodeCalib)) {
throw new IllegalArgumentException("Calibrated Yule model requires a calibration to be specified for node, " +
nodeCalib.getName() + ".");
}
}
}
writer.writeCloseTag(SpeciationLikelihoodParser.SPECIATION_LIKELIHOOD);
break;
// AR - removed this special case for Logistic - it causes terrible problems.
// Need to put informative priors on Logistic parameters.
// case LOGISTIC:
// writer.writeComment("Generate a boolean likelihood for Coalescent: Logistic Growth");
// writer.writeOpenTag(
// BooleanLikelihoodParser.BOOLEAN_LIKELIHOOD,
// new Attribute[]{new Attribute.Default<String>(XMLParser.ID, modelPrefix + "booleanLikelihood1")}
// );
// writer.writeOpenTag(
// TestStatisticParser.TEST_STATISTIC,
// new Attribute[]{
// new Attribute.Default<String>(XMLParser.ID, "test1"),
// new Attribute.Default<String>(Statistic.NAME, "test1")
// }
// );
// writer.writeIDref(ParameterParser.PARAMETER, priorPrefix + "logistic.t50");
// writer.writeOpenTag("lessThan");
// writer.writeIDref(ParameterParser.PARAMETER, modelPrefix + "treeModel.rootHeight");
// writer.writeCloseTag("lessThan");
// writer.writeCloseTag(TestStatisticParser.TEST_STATISTIC);
// writer.writeCloseTag(BooleanLikelihoodParser.BOOLEAN_LIKELIHOOD);
//
// writer.writeOpenTag(
// CoalescentLikelihoodParser.COALESCENT_LIKELIHOOD,
// new Attribute[]{new Attribute.Default<String>(XMLParser.ID, modelPrefix + COALESCENT)}
// );
// writer.writeOpenTag(CoalescentLikelihoodParser.MODEL);
// writeNodeHeightPriorModelRef(prior, writer);
// writer.writeCloseTag(CoalescentLikelihoodParser.MODEL);
// writer.writeOpenTag(CoalescentLikelihoodParser.POPULATION_TREE);
// writer.writeIDref(TreeModel.TREE_MODEL, modelPrefix + TreeModel.TREE_MODEL);
// writer.writeCloseTag(CoalescentLikelihoodParser.POPULATION_TREE);
// writer.writeCloseTag(CoalescentLikelihoodParser.COALESCENT_LIKELIHOOD);
//
// break;
case SKYLINE:
// generate a Bayesian skyline plot
writer.writeComment("Generate a generalizedSkyLineLikelihood for Bayesian Skyline");
writer.writeOpenTag(
BayesianSkylineLikelihoodParser.SKYLINE_LIKELIHOOD,
new Attribute[]{
new Attribute.Default<String>(XMLParser.ID, modelPrefix + "skyline"),
new Attribute.Default<String>("linear",
prior.getSkylineModel() == TreePriorParameterizationType.LINEAR_SKYLINE ? "true" : "false")
}
);
// write pop size socket
writer.writeOpenTag(BayesianSkylineLikelihoodParser.POPULATION_SIZES);
if (prior.getSkylineModel() == TreePriorParameterizationType.LINEAR_SKYLINE) {
writeParameter(prior.getParameter("skyline.popSize"), prior.getSkylineGroupCount() + 1, writer);
} else {
writeParameter(prior.getParameter("skyline.popSize"), prior.getSkylineGroupCount(), writer);
}
writer.writeCloseTag(BayesianSkylineLikelihoodParser.POPULATION_SIZES);
// write group size socket
writer.writeOpenTag(BayesianSkylineLikelihoodParser.GROUP_SIZES);
writeParameter(prior.getParameter("skyline.groupSize"), prior.getSkylineGroupCount(), writer);
writer.writeCloseTag(BayesianSkylineLikelihoodParser.GROUP_SIZES);
writer.writeOpenTag(CoalescentLikelihoodParser.POPULATION_TREE);
writer.writeIDref(TreeModel.TREE_MODEL, modelPrefix + TreeModel.TREE_MODEL);
writer.writeCloseTag(CoalescentLikelihoodParser.POPULATION_TREE);
writer.writeCloseTag(BayesianSkylineLikelihoodParser.SKYLINE_LIKELIHOOD);
writer.writeText("");
writeExponentialMarkovLikelihood(prior, writer);
break;
case EXTENDED_SKYLINE:
// different format
break;
case GMRF_SKYRIDE:
writer.writeComment("Generate a gmrfSkyrideLikelihood for GMRF Bayesian Skyride process");
writer.writeOpenTag(
GMRFSkyrideLikelihoodParser.SKYLINE_LIKELIHOOD,
new Attribute[]{
new Attribute.Default<String>(XMLParser.ID, modelPrefix + "skyride"),
new Attribute.Default<String>(GMRFSkyrideLikelihoodParser.TIME_AWARE_SMOOTHING,
prior.getSkyrideSmoothing() == TreePriorParameterizationType.TIME_AWARE_SKYRIDE ? "true" : "false"),
new Attribute.Default<String>(GMRFSkyrideLikelihoodParser.RANDOMIZE_TREE,
//TODO For GMRF, tree model/tree prior combination not implemented by BEAST yet. The validation is in BeastGenerator.checkOptions()
options.getPartitionTreeModels(prior).get(0).getStartingTreeType() == StartingTreeType.UPGMA ? "true" : "false"),
}
);
int skyrideIntervalCount = options.taxonList.getTaxonCount() - 1;
writer.writeOpenTag(GMRFSkyrideLikelihoodParser.POPULATION_PARAMETER);
writer.writeComment("skyride.logPopSize is in log units unlike other popSize");
writeParameter(prior.getParameter("skyride.logPopSize"), skyrideIntervalCount, writer);
writer.writeCloseTag(GMRFSkyrideLikelihoodParser.POPULATION_PARAMETER);
writer.writeOpenTag(GMRFSkyrideLikelihoodParser.GROUP_SIZES);
writeParameter(prior.getParameter("skyride.groupSize"), skyrideIntervalCount, writer);
writer.writeCloseTag(GMRFSkyrideLikelihoodParser.GROUP_SIZES);
writer.writeOpenTag(GMRFSkyrideLikelihoodParser.PRECISION_PARAMETER);
writeParameter(prior.getParameter("skyride.precision"), 1, writer);
writer.writeCloseTag(GMRFSkyrideLikelihoodParser.PRECISION_PARAMETER);
writer.writeOpenTag(GMRFSkyrideLikelihoodParser.POPULATION_TREE);
writer.writeIDref(TreeModel.TREE_MODEL, modelPrefix + TreeModel.TREE_MODEL);
writer.writeCloseTag(GMRFSkyrideLikelihoodParser.POPULATION_TREE);
writer.writeCloseTag(GMRFSkyrideLikelihoodParser.SKYLINE_LIKELIHOOD);
break;
case SKYGRID:
break;
case SPECIES_YULE:
case SPECIES_YULE_CALIBRATION:
case SPECIES_BIRTH_DEATH:
break;
default:
// generate a coalescent process
writer.writeComment("Generate a coalescent likelihood");
writer.writeOpenTag(
CoalescentLikelihoodParser.COALESCENT_LIKELIHOOD,
new Attribute[]{new Attribute.Default<String>(XMLParser.ID, modelPrefix + COALESCENT)}
);
writer.writeOpenTag(CoalescentLikelihoodParser.MODEL);
writeNodeHeightPriorModelRef(prior, writer);
writer.writeCloseTag(CoalescentLikelihoodParser.MODEL);
writer.writeOpenTag(CoalescentLikelihoodParser.POPULATION_TREE);
writer.writeIDref(TreeModel.TREE_MODEL, modelPrefix + TreeModel.TREE_MODEL);
writer.writeCloseTag(CoalescentLikelihoodParser.POPULATION_TREE);
writer.writeCloseTag(CoalescentLikelihoodParser.COALESCENT_LIKELIHOOD);
}
}
void writeNodeHeightPriorModelRef(PartitionTreePrior prior, XMLWriter writer) {
TreePriorType treePrior = prior.getNodeHeightPrior();
String priorPrefix = prior.getPrefix();
switch (treePrior) {
case CONSTANT:
case SPECIES_YULE:
case SPECIES_BIRTH_DEATH:
case SPECIES_YULE_CALIBRATION:
writer.writeIDref(ConstantPopulationModelParser.CONSTANT_POPULATION_MODEL, priorPrefix + "constant");
break;
case EXPONENTIAL:
writer.writeIDref(ExponentialGrowthModelParser.EXPONENTIAL_GROWTH_MODEL, priorPrefix + "exponential");
break;
case LOGISTIC:
writer.writeIDref(LogisticGrowthModelParser.LOGISTIC_GROWTH_MODEL, priorPrefix + "logistic");
break;
case EXPANSION:
writer.writeIDref(ExpansionModelParser.EXPANSION_MODEL, priorPrefix + "expansion");
break;
case SKYLINE:
writer.writeIDref(BayesianSkylineLikelihoodParser.SKYLINE_LIKELIHOOD, priorPrefix + "skyline");
break;
case GMRF_SKYRIDE:
writer.writeIDref(GMRFSkyrideLikelihoodParser.SKYLINE_LIKELIHOOD, priorPrefix + "skyride");
break;
case SKYGRID:
writer.writeIDref(GMRFSkyrideLikelihoodParser.SKYGRID_LIKELIHOOD, priorPrefix + "skygrid");
break;
case YULE:
case YULE_CALIBRATION:
writer.writeIDref(YuleModelParser.YULE_MODEL, priorPrefix + YuleModelParser.YULE);
break;
case BIRTH_DEATH:
case BIRTH_DEATH_INCOMPLETE_SAMPLING:
writer.writeIDref(BirthDeathModelParser.BIRTH_DEATH_MODEL, priorPrefix + BirthDeathModelParser.BIRTH_DEATH);
break;
case BIRTH_DEATH_SERIAL_SAMPLING:
writer.writeIDref(BirthDeathSerialSamplingModelParser.BIRTH_DEATH_SERIAL_MODEL,
priorPrefix + BirthDeathSerialSamplingModelParser.BDSS);
break;
case BIRTH_DEATH_BASIC_REPRODUCTIVE_NUMBER:
writer.writeIDref(BirthDeathEpidemiologyModelParser.BIRTH_DEATH_EPIDEMIOLOGY,
priorPrefix + BirthDeathEpidemiologyModelParser.BIRTH_DEATH_EPIDEMIOLOGY);
break;
default:
throw new IllegalArgumentException("No tree prior has been specified so cannot refer to it");
}
}
void writeMultiLociTreePriors(PartitionTreePrior prior, XMLWriter writer) {
if (prior.getNodeHeightPrior() == TreePriorType.SKYGRID) {
setModelPrefix(prior.getPrefix());
writer.writeComment("Generate a gmrfSkyGridLikelihood for the Bayesian SkyGrid process");
writer.writeOpenTag(
GMRFSkyrideLikelihoodParser.SKYGRID_LIKELIHOOD,
new Attribute[]{
new Attribute.Default<String>(XMLParser.ID, modelPrefix + "skygrid"),
}
);
int skyGridIntervalCount = prior.getSkyGridCount();
double skyGridInterval = prior.getSkyGridInterval();
writer.writeOpenTag(GMRFSkyrideLikelihoodParser.POPULATION_PARAMETER);
writer.writeComment("skygrid.logPopSize is in log units unlike other popSize");
writeParameter(prior.getParameter("skygrid.logPopSize"), skyGridIntervalCount, writer);
writer.writeCloseTag(GMRFSkyrideLikelihoodParser.POPULATION_PARAMETER);
writer.writeOpenTag(GMRFSkyrideLikelihoodParser.PRECISION_PARAMETER);
writeParameter(prior.getParameter("skygrid.precision"), 1, writer);
writer.writeCloseTag(GMRFSkyrideLikelihoodParser.PRECISION_PARAMETER);
writer.writeOpenTag(GMRFSkyrideLikelihoodParser.NUM_GRID_POINTS);
Parameter numGridPoint = prior.getParameter("skygrid.numGridPoints");
numGridPoint.setInitial(skyGridIntervalCount - 1);
writeParameter(numGridPoint, 1, writer);
writer.writeCloseTag(GMRFSkyrideLikelihoodParser.NUM_GRID_POINTS);
writer.writeOpenTag(GMRFSkyrideLikelihoodParser.CUT_OFF);
Parameter cutOff = prior.getParameter("skygrid.cutOff");
cutOff.setInitial(skyGridInterval);
writeParameter(cutOff, 1, writer);
writer.writeCloseTag(GMRFSkyrideLikelihoodParser.CUT_OFF);
writer.writeOpenTag(GMRFSkyrideLikelihoodParser.POPULATION_TREE);
// TODO Add all linked trees
if (options.isShareSameTreePrior()) {
for (PartitionTreeModel thisModel : options.getPartitionTreeModels()) {
writer.writeIDref(TreeModel.TREE_MODEL, thisModel.getPrefix() + TreeModel.TREE_MODEL);
}
} else {
writer.writeIDref(TreeModel.TREE_MODEL, options.getPartitionTreeModels(prior).get(0).getPrefix() + TreeModel.TREE_MODEL);
}
writer.writeCloseTag(GMRFSkyrideLikelihoodParser.POPULATION_TREE);
writer.writeCloseTag(GMRFSkyrideLikelihoodParser.SKYGRID_LIKELIHOOD);
} else if (prior.getNodeHeightPrior() == TreePriorType.EXTENDED_SKYLINE) {
setModelPrefix(prior.getPrefix());
final String tagName = VariableDemographicModelParser.MODEL_NAME;
writer.writeComment("Generate a variableDemographic for extended Bayesian skyline process");
writer.writeOpenTag(tagName, new Attribute[]{
new Attribute.Default<String>(XMLParser.ID, modelPrefix + VariableDemographicModelParser.demoElementName),
new Attribute.Default<String>(VariableDemographicModelParser.TYPE, prior.getExtendedSkylineModel().toString()),
// use midpoint by default (todo) would be nice to have a user 'tickable' option
new Attribute.Default<String>(VariableDemographicModelParser.USE_MIDPOINTS, "true")
}
);
Parameter popSize = prior.getParameter(VariableDemographicModelParser.demoElementName + ".popSize");
Parameter populationMean = prior.getParameter(VariableDemographicModelParser.demoElementName + ".populationMean");
popSize.setInitial(populationMean.getInitial());
writer.writeOpenTag(VariableDemographicModelParser.POPULATION_SIZES);
writer.writeComment("popSize value = populationMean value");
writer.writeTag(ParameterParser.PARAMETER,
new Attribute[]{
new Attribute.Default<String>(XMLParser.ID, modelPrefix + VariableDemographicModelParser.demoElementName + ".popSize"),
new Attribute.Default<String>(ParameterParser.VALUE, Double.toString(popSize.getInitial()))}, true);
// writeParameter(popSize, -1, writer);
writer.writeCloseTag(VariableDemographicModelParser.POPULATION_SIZES);
// Parameter indicators = prior.getParameter(VariableDemographicModelParser.demoElementName + ".indicators");
writer.writeOpenTag(VariableDemographicModelParser.INDICATOR_PARAMETER);
writer.writeTag(ParameterParser.PARAMETER,
new Attribute[]{
new Attribute.Default<String>(XMLParser.ID, modelPrefix + VariableDemographicModelParser.demoElementName + ".indicators"),
new Attribute.Default<String>(ParameterParser.VALUE, Double.toString(0.0))}, true); // also 0.0
// writeParameter(prior.getParameter(VariableDemographicModelParser.demoElementName + ".indicators"), -1, writer); // not need dimension
writer.writeCloseTag(VariableDemographicModelParser.INDICATOR_PARAMETER);
writer.writeOpenTag(VariableDemographicModelParser.POPULATION_TREES);
if (options.isShareSameTreePrior()) {
for (PartitionTreeModel model : options.getPartitionTreeModels()) {
writer.writeOpenTag(VariableDemographicModelParser.POP_TREE, new Attribute[]{
new Attribute.Default<String>(SpeciesBindingsParser.PLOIDY, Double.toString(model.getPloidyType().getValue()))
}
);
writer.writeIDref(TreeModel.TREE_MODEL, model.getPrefix() + TreeModel.TREE_MODEL);
writer.writeCloseTag(VariableDemographicModelParser.POP_TREE);
}
} else {//TODO correct for not sharing same prior?
writer.writeOpenTag(VariableDemographicModelParser.POP_TREE, new Attribute[]{
new Attribute.Default<String>(SpeciesBindingsParser.PLOIDY, Double.toString(options.getPartitionTreeModels(prior).get(0).getPloidyType().getValue()))
}
);
writer.writeIDref(TreeModel.TREE_MODEL, options.getPartitionTreeModels(prior).get(0).getPrefix() + TreeModel.TREE_MODEL);
writer.writeCloseTag(VariableDemographicModelParser.POP_TREE);
}
writer.writeCloseTag(VariableDemographicModelParser.POPULATION_TREES);
writer.writeCloseTag(tagName);
writer.writeOpenTag(CoalescentLikelihoodParser.COALESCENT_LIKELIHOOD, new Attribute.Default<String>(XMLParser.ID, modelPrefix + COALESCENT));
writer.writeOpenTag(CoalescentLikelihoodParser.MODEL);
writer.writeIDref(tagName, modelPrefix + VariableDemographicModelParser.demoElementName);
writer.writeCloseTag(CoalescentLikelihoodParser.MODEL);
writer.writeComment("Take population Tree from demographic");
writer.writeCloseTag(CoalescentLikelihoodParser.COALESCENT_LIKELIHOOD);
writer.writeOpenTag(SumStatisticParser.SUM_STATISTIC,
new Attribute[]{
new Attribute.Default<String>(XMLParser.ID, modelPrefix + VariableDemographicModelParser.demoElementName + ".populationSizeChanges"),
new Attribute.Default<String>("elementwise", "true")
});
writer.writeIDref(ParameterParser.PARAMETER, modelPrefix + VariableDemographicModelParser.demoElementName + ".indicators");
writer.writeCloseTag(SumStatisticParser.SUM_STATISTIC);
writer.writeOpenTag(ExponentialDistributionModel.EXPONENTIAL_DISTRIBUTION_MODEL,
new Attribute[]{
new Attribute.Default<String>(XMLParser.ID, modelPrefix + VariableDemographicModelParser.demoElementName + ".populationMeanDist")
//,new Attribute.Default<String>("elementwise", "true")
});
writer.writeOpenTag(DistributionModelParser.MEAN);
writer.writeComment("prefer populationMean value = 1");
populationMean = prior.getParameter(VariableDemographicModelParser.demoElementName + ".populationMean");
writer.writeTag(ParameterParser.PARAMETER,
new Attribute[]{
new Attribute.Default<String>(XMLParser.ID, modelPrefix + VariableDemographicModelParser.demoElementName + ".populationMean"),
new Attribute.Default<String>(ParameterParser.VALUE, Double.toString(populationMean.getInitial()))}, true);
writer.writeCloseTag(DistributionModelParser.MEAN);
writer.writeCloseTag(ExponentialDistributionModel.EXPONENTIAL_DISTRIBUTION_MODEL);
}
}
void writeParameterLog(PartitionTreePrior prior, XMLWriter writer) {
setModelPrefix(prior.getPrefix());
switch (prior.getNodeHeightPrior()) {
case CONSTANT:
writeParameterRef(modelPrefix + "constant.popSize", writer);
break;
case EXPONENTIAL:
writeParameterRef(modelPrefix + "exponential.popSize", writer);
if (prior.getParameterization() == TreePriorParameterizationType.GROWTH_RATE) {
writeParameterRef(modelPrefix + "exponential.growthRate", writer);
} else {
writeParameterRef(modelPrefix + "exponential.doublingTime", writer);
}
break;
case LOGISTIC:
writeParameterRef(modelPrefix + "logistic.popSize", writer);
if (prior.getParameterization() == TreePriorParameterizationType.GROWTH_RATE) {
writeParameterRef(modelPrefix + "logistic.growthRate", writer);
} else {
writeParameterRef(modelPrefix + "logistic.doublingTime", writer);
}
writeParameterRef(modelPrefix + "logistic.t50", writer);
break;
case EXPANSION:
writeParameterRef(modelPrefix + "expansion.popSize", writer);
if (prior.getParameterization() == TreePriorParameterizationType.GROWTH_RATE) {
writeParameterRef(modelPrefix + "expansion.growthRate", writer);
} else {
writeParameterRef(modelPrefix + "expansion.doublingTime", writer);
}
writeParameterRef(modelPrefix + "expansion.ancestralProportion", writer);
break;
case SKYLINE:
writeParameterRef(modelPrefix + "skyline.popSize", writer);
writeParameterRef(modelPrefix + "skyline.groupSize", writer);
break;
case EXTENDED_SKYLINE:
writer.writeIDref(SumStatisticParser.SUM_STATISTIC, "demographic.populationSizeChanges");
writeParameterRef(modelPrefix + "demographic.populationMean", writer);
writeParameterRef(modelPrefix + "demographic.popSize", writer);
writeParameterRef(modelPrefix + "demographic.indicators", writer);
break;
case SKYGRID:
writeParameterRef(modelPrefix + "skygrid.precision", writer);
writeParameterRef(modelPrefix + "skygrid.logPopSize", writer);
writeParameterRef(modelPrefix + "skygrid.cutOff", writer);
// writeParameterRef(modelPrefix + "skygrid.groupSize", writer);
break;
case GMRF_SKYRIDE:
writeParameterRef(modelPrefix + "skyride.precision", writer);
writeParameterRef(modelPrefix + "skyride.logPopSize", writer);
writeParameterRef(modelPrefix + "skyride.groupSize", writer);
break;
case YULE:
case YULE_CALIBRATION:
writeParameterRef(modelPrefix + "yule.birthRate", writer);
break;
case BIRTH_DEATH:
case BIRTH_DEATH_INCOMPLETE_SAMPLING:
writeParameterRef(modelPrefix + BirthDeathModelParser.MEAN_GROWTH_RATE_PARAM_NAME, writer);
writeParameterRef(modelPrefix + BirthDeathModelParser.RELATIVE_DEATH_RATE_PARAM_NAME, writer);
if (prior.getNodeHeightPrior() == TreePriorType.BIRTH_DEATH_INCOMPLETE_SAMPLING)
writeParameterRef(modelPrefix + BirthDeathModelParser.BIRTH_DEATH + "."
+ BirthDeathModelParser.SAMPLE_PROB, writer);
break;
case BIRTH_DEATH_SERIAL_SAMPLING:
writeParameterRef(modelPrefix + BirthDeathSerialSamplingModelParser.BDSS + "."
+ BirthDeathSerialSamplingModelParser.LAMBDA, writer);
writeParameterRef(modelPrefix + BirthDeathSerialSamplingModelParser.BDSS + "."
+ BirthDeathSerialSamplingModelParser.RELATIVE_MU, writer);
//Issue 656
// writeParameterRef(modelPrefix + BirthDeathSerialSamplingModelParser.BDSS + "."
// + BirthDeathSerialSamplingModelParser.SAMPLE_PROBABILITY, writer);
writeParameterRef(modelPrefix + BirthDeathSerialSamplingModelParser.BDSS + "."
+ BirthDeathSerialSamplingModelParser.PSI, writer);
writeParameterRef(modelPrefix + BirthDeathSerialSamplingModelParser.BDSS + "."
+ BirthDeathSerialSamplingModelParser.ORIGIN, writer);
break;
case BIRTH_DEATH_BASIC_REPRODUCTIVE_NUMBER:
writeParameterRef(modelPrefix + BirthDeathEpidemiologyModelParser.R0, writer);
writeParameterRef(modelPrefix + BirthDeathEpidemiologyModelParser.RECOVERY_RATE, writer);
writeParameterRef(modelPrefix + BirthDeathEpidemiologyModelParser.SAMPLING_PROBABILITY, writer);
writeParameterRef(modelPrefix + BirthDeathEpidemiologyModelParser.ORIGIN, writer);
case SPECIES_YULE:
case SPECIES_BIRTH_DEATH:
case SPECIES_YULE_CALIBRATION:
break;
default:
throw new IllegalArgumentException("No tree prior has been specified so cannot refer to it");
}
}
void writeEBSPAnalysisToCSVfile(PartitionTreePrior prior, XMLWriter writer) {
setModelPrefix(prior.getPrefix());
String logFileName = options.logFileName;
if (prior.getNodeHeightPrior() == TreePriorType.EXTENDED_SKYLINE) {
writer.writeOpenTag(EBSPAnalysisParser.VD_ANALYSIS, new Attribute[]{
new Attribute.Default<String>(XMLParser.ID, modelPrefix + "demographic.analysis"),
new Attribute.Default<Double>(EBSPAnalysisParser.BURN_IN, 0.1),
new Attribute.Default<Boolean>(VariableDemographicModelParser.USE_MIDPOINTS, true)}
);
writer.writeOpenTag(EBSPAnalysisParser.LOG_FILE_NAME);
writer.writeText(logFileName);
writer.writeCloseTag(EBSPAnalysisParser.LOG_FILE_NAME);
writer.writeOpenTag(EBSPAnalysisParser.TREE_FILE_NAMES);
for (String treeFN : options.treeFileName) {
writer.writeOpenTag(EBSPAnalysisParser.TREE_LOG);
writer.writeText(treeFN);
writer.writeCloseTag(EBSPAnalysisParser.TREE_LOG);
}
writer.writeCloseTag(EBSPAnalysisParser.TREE_FILE_NAMES);
writer.writeOpenTag(EBSPAnalysisParser.MODEL_TYPE);
writer.writeText(prior.getExtendedSkylineModel().toString());
writer.writeCloseTag(EBSPAnalysisParser.MODEL_TYPE);
writer.writeOpenTag(EBSPAnalysisParser.POPULATION_FIRST_COLUMN);
writer.writeText(VariableDemographicModelParser.demoElementName + ".popSize1");
writer.writeCloseTag(EBSPAnalysisParser.POPULATION_FIRST_COLUMN);
writer.writeOpenTag(EBSPAnalysisParser.INDICATORS_FIRST_COLUMN);
writer.writeText(VariableDemographicModelParser.demoElementName + ".indicators1");
writer.writeCloseTag(EBSPAnalysisParser.INDICATORS_FIRST_COLUMN);
writer.writeCloseTag(EBSPAnalysisParser.VD_ANALYSIS);
writer.writeOpenTag(CSVExporterParser.CSV_EXPORT,
new Attribute[]{
new Attribute.Default<String>(CSVExporterParser.FILE_NAME,
logFileName.subSequence(0, logFileName.length() - 4) + ".csv"), //.log
new Attribute.Default<String>(CSVExporterParser.SEPARATOR, ",")
});
writer.writeOpenTag(CSVExporterParser.COLUMNS);
writer.writeIDref(EBSPAnalysisParser.VD_ANALYSIS, modelPrefix + "demographic.analysis");
writer.writeCloseTag(CSVExporterParser.COLUMNS);
writer.writeCloseTag(CSVExporterParser.CSV_EXPORT);
}
}
private void writeExponentialMarkovLikelihood(PartitionTreePrior prior, XMLWriter writer) {
writer.writeOpenTag(
ExponentialMarkovModel.EXPONENTIAL_MARKOV_MODEL,
new Attribute[]{new Attribute.Default<String>(XMLParser.ID, modelPrefix + "eml1"),
new Attribute.Default<String>("jeffreys", "true")}
);
writeParameterRef(ExponentialMarkovModelParser.CHAIN_PARAMETER, prior.getPrefix() + "skyline.popSize", writer);
writer.writeCloseTag(ExponentialMarkovModel.EXPONENTIAL_MARKOV_MODEL);
}
public static void writePriorLikelihoodReferenceLog(PartitionTreePrior prior, PartitionTreeModel model, XMLWriter writer) {
//tree model prefix
String modelPrefix = model.getPrefix(); // only has prefix, if (options.getPartitionTreePriors().size() > 1)
switch (prior.getNodeHeightPrior()) {
case YULE:
case YULE_CALIBRATION:
case BIRTH_DEATH:
case BIRTH_DEATH_INCOMPLETE_SAMPLING:
case BIRTH_DEATH_SERIAL_SAMPLING:
case BIRTH_DEATH_BASIC_REPRODUCTIVE_NUMBER:
writer.writeIDref(SpeciationLikelihoodParser.SPECIATION_LIKELIHOOD, modelPrefix + "speciation");
break;
case SKYLINE:
writer.writeIDref(BayesianSkylineLikelihoodParser.SKYLINE_LIKELIHOOD, modelPrefix + "skyline");
// writer.writeIDref(ExponentialMarkovModel.EXPONENTIAL_MARKOV_MODEL, modelPrefix + "eml1");
break;
case GMRF_SKYRIDE:
writer.writeIDref(GMRFSkyrideLikelihoodParser.SKYLINE_LIKELIHOOD, modelPrefix + "skyride");
break;
case SKYGRID:
writer.writeIDref(GMRFSkyrideLikelihoodParser.SKYLINE_LIKELIHOOD, modelPrefix + "skygrid");
// only 1 coalescent, so write it separately after this method
break;
case LOGISTIC:
// writer.writeIDref(BooleanLikelihoodParser.BOOLEAN_LIKELIHOOD, modelPrefix + "booleanLikelihood1");
writer.writeIDref(CoalescentLikelihoodParser.COALESCENT_LIKELIHOOD, modelPrefix + COALESCENT);
break;
case EXTENDED_SKYLINE:
// only 1 coalescent, so write it separately after this method
case SPECIES_YULE:
case SPECIES_YULE_CALIBRATION:
case SPECIES_BIRTH_DEATH:
// do not need
break;
default:
writer.writeIDref(CoalescentLikelihoodParser.COALESCENT_LIKELIHOOD, modelPrefix + COALESCENT);
}
}
// id is written in writePriorLikelihood (PartitionTreePrior prior, PartitionTreeModel model, XMLWriter writer)
public void writePriorLikelihoodReference(PartitionTreePrior prior, PartitionTreeModel model, XMLWriter writer) {
//tree model prefix
setModelPrefix(model.getPrefix()); // only has prefix, if (options.getPartitionTreePriors().size() > 1)
switch (prior.getNodeHeightPrior()) {
case YULE:
case YULE_CALIBRATION:
case BIRTH_DEATH:
case BIRTH_DEATH_INCOMPLETE_SAMPLING:
case BIRTH_DEATH_SERIAL_SAMPLING:
case BIRTH_DEATH_BASIC_REPRODUCTIVE_NUMBER:
writer.writeIDref(SpeciationLikelihoodParser.SPECIATION_LIKELIHOOD, modelPrefix + "speciation");
break;
case SKYLINE:
writer.writeIDref(BayesianSkylineLikelihoodParser.SKYLINE_LIKELIHOOD, modelPrefix + "skyline");
writer.writeIDref(ExponentialMarkovModel.EXPONENTIAL_MARKOV_MODEL, modelPrefix + "eml1");
break;
case GMRF_SKYRIDE:
writer.writeIDref(GMRFSkyrideLikelihoodParser.SKYLINE_LIKELIHOOD, modelPrefix + "skyride");
break;
case SKYGRID:
// writer.writeIDref(GMRFSkyrideLikelihoodParser.SKYLINE_LIKELIHOOD, modelPrefix + "skygrid");
// only 1 coalescent, so write it separately after this method
break;
// case LOGISTIC:
// writer.writeIDref(BooleanLikelihoodParser.BOOLEAN_LIKELIHOOD, modelPrefix + "booleanLikelihood1");
// writer.writeIDref(CoalescentLikelihoodParser.COALESCENT_LIKELIHOOD, modelPrefix + COALESCENT);
// break;
case EXTENDED_SKYLINE:
// only 1 coalescent, so write it in writeEBSPVariableDemographicReference
case SPECIES_YULE:
case SPECIES_YULE_CALIBRATION:
case SPECIES_BIRTH_DEATH:
// do not need
break;
default:
writer.writeIDref(CoalescentLikelihoodParser.COALESCENT_LIKELIHOOD, modelPrefix + COALESCENT);
}
}
public void writeMultiLociLikelihoodReference(PartitionTreePrior prior, XMLWriter writer) {
setModelPrefix(prior.getPrefix());
//TODO: make suitable for *BEAST
if (prior.getNodeHeightPrior() == TreePriorType.EXTENDED_SKYLINE) {
writer.writeIDref(CoalescentLikelihoodParser.COALESCENT_LIKELIHOOD, modelPrefix + COALESCENT); // only 1 coalescent
writer.writeOpenTag(MixedDistributionLikelihoodParser.DISTRIBUTION_LIKELIHOOD);
writer.writeOpenTag(MixedDistributionLikelihoodParser.DISTRIBUTION0);
writer.writeIDref(ExponentialDistributionModel.EXPONENTIAL_DISTRIBUTION_MODEL, modelPrefix + "demographic.populationMeanDist");
writer.writeCloseTag(MixedDistributionLikelihoodParser.DISTRIBUTION0);
writer.writeOpenTag(MixedDistributionLikelihoodParser.DISTRIBUTION1);
writer.writeIDref(ExponentialDistributionModel.EXPONENTIAL_DISTRIBUTION_MODEL, modelPrefix + "demographic.populationMeanDist");
writer.writeCloseTag(MixedDistributionLikelihoodParser.DISTRIBUTION1);
writeParameterRef(MixedDistributionLikelihoodParser.DATA, modelPrefix + "demographic.popSize", writer);
writeParameterRef(MixedDistributionLikelihoodParser.INDICATORS, modelPrefix + "demographic.indicators", writer);
writer.writeCloseTag(MixedDistributionLikelihoodParser.DISTRIBUTION_LIKELIHOOD);
} else if (prior.getNodeHeightPrior() == TreePriorType.SKYGRID) {
writer.writeIDref(GMRFSkyrideLikelihoodParser.SKYGRID_LIKELIHOOD, modelPrefix + "skygrid");
}
}
}