/**
* Combines two recalibration reports by adding all observations and errors
*
* <p>Note: This method DOES NOT recalculate the empirical qualities and quantized qualities. You
* have to recalculate them after combining. The reason for not calculating it is because this
* function is intended for combining a series of recalibration reports, and it only makes sense
* to calculate the empirical qualities and quantized qualities after all the recalibration
* reports have been combined. Having the user recalculate when appropriate, makes this method
* faster
*
* <p>Note2: The empirical quality reported, however, is recalculated given its simplicity.
*
* @param other the recalibration report to combine with this one
*/
public void combine(final RecalibrationReport other) {
for (int tableIndex = 0; tableIndex < recalibrationTables.numTables(); tableIndex++) {
final NestedIntegerArray<RecalDatum> myTable = recalibrationTables.getTable(tableIndex);
final NestedIntegerArray<RecalDatum> otherTable =
other.recalibrationTables.getTable(tableIndex);
RecalUtils.combineTables(myTable, otherTable);
}
}
public RecalibrationReport(final File recalFile, final SortedSet<String> allReadGroups) {
final GATKReport report = new GATKReport(recalFile);
argumentTable = report.getTable(RecalUtils.ARGUMENT_REPORT_TABLE_TITLE);
RAC = initializeArgumentCollectionTable(argumentTable);
GATKReportTable quantizedTable = report.getTable(RecalUtils.QUANTIZED_REPORT_TABLE_TITLE);
quantizationInfo = initializeQuantizationTable(quantizedTable);
Pair<ArrayList<Covariate>, ArrayList<Covariate>> covariates =
RecalUtils.initializeCovariates(RAC); // initialize the required and optional covariates
ArrayList<Covariate> requiredCovariates = covariates.getFirst();
ArrayList<Covariate> optionalCovariates = covariates.getSecond();
requestedCovariates = new Covariate[requiredCovariates.size() + optionalCovariates.size()];
optionalCovariateIndexes = new HashMap<String, Integer>(optionalCovariates.size());
int covariateIndex = 0;
for (final Covariate covariate : requiredCovariates)
requestedCovariates[covariateIndex++] = covariate;
for (final Covariate covariate : optionalCovariates) {
requestedCovariates[covariateIndex] = covariate;
final String covariateName =
covariate
.getClass()
.getSimpleName()
.split("Covariate")[
0]; // get the name of the covariate (without the "covariate" part of it) so we can
// match with the GATKReport
optionalCovariateIndexes.put(covariateName, covariateIndex - 2);
covariateIndex++;
}
for (Covariate cov : requestedCovariates)
cov.initialize(
RAC); // initialize any covariate member variables using the shared argument collection
recalibrationTables = new RecalibrationTables(requestedCovariates, allReadGroups.size());
initializeReadGroupCovariates(allReadGroups);
parseReadGroupTable(
report.getTable(RecalUtils.READGROUP_REPORT_TABLE_TITLE),
recalibrationTables.getReadGroupTable());
parseQualityScoreTable(
report.getTable(RecalUtils.QUALITY_SCORE_REPORT_TABLE_TITLE),
recalibrationTables.getQualityScoreTable());
parseAllCovariatesTable(
report.getTable(RecalUtils.ALL_COVARIATES_REPORT_TABLE_TITLE), recalibrationTables);
}
/**
* Compiles the list of keys for the Covariates table and uses the shared parsing utility to
* produce the actual table
*
* @param reportTable the GATKReport table containing data for this table
* @param recalibrationTables the recalibration tables \
*/
private void parseAllCovariatesTable(
final GATKReportTable reportTable, final RecalibrationTables recalibrationTables) {
for (int i = 0; i < reportTable.getNumRows(); i++) {
final Object rg = reportTable.get(i, RecalUtils.READGROUP_COLUMN_NAME);
tempCOVarray[0] = requestedCovariates[0].keyFromValue(rg);
final Object qual = reportTable.get(i, RecalUtils.QUALITY_SCORE_COLUMN_NAME);
tempCOVarray[1] = requestedCovariates[1].keyFromValue(qual);
final String covName = (String) reportTable.get(i, RecalUtils.COVARIATE_NAME_COLUMN_NAME);
final int covIndex = optionalCovariateIndexes.get(covName);
final Object covValue = reportTable.get(i, RecalUtils.COVARIATE_VALUE_COLUMN_NAME);
tempCOVarray[2] =
requestedCovariates[
RecalibrationTables.TableType.OPTIONAL_COVARIATE_TABLES_START.ordinal() + covIndex]
.keyFromValue(covValue);
final EventType event =
EventType.eventFrom((String) reportTable.get(i, RecalUtils.EVENT_TYPE_COLUMN_NAME));
tempCOVarray[3] = event.ordinal();
recalibrationTables
.getTable(
RecalibrationTables.TableType.OPTIONAL_COVARIATE_TABLES_START.ordinal() + covIndex)
.put(getRecalDatum(reportTable, i, false), tempCOVarray);
}
}
/** @return true if the report has no data */
public boolean isEmpty() {
return recalibrationTables.isEmpty();
}