/**
* Since the segmentations are disjoint, the total set of differentially expressed segments is the
* union of the sets of differentially expressed segments in each segmentation.
*
* @param c1
* @param c2
* @return
*/
public SortedSet<DifferentialKey> differentialRegions(int c1, int c2) {
SortedSet<DifferentialKey> keys = new TreeSet<DifferentialKey>();
for (InputSegmentation iseg : segmentations) {
keys.addAll(differentialRegions(iseg, c1, c2));
}
return keys;
}
public static void main(String[] args) throws Exception {
// TODO Auto-generated method stub
Date startTime = new Date();
String fname = "p079_keylog.txt";
FileReader fr = new FileReader(fname);
BufferedReader br = new BufferedReader(fr);
SortedSet<String> slines = new TreeSet<String>();
String line = "";
while ((line = br.readLine()) != null) {
slines.add(line);
}
String start = slines.first().split("")[0];
String ans = "";
for (String l : slines) {
// System.out.println(l);
for (String l2 : slines) {
if (l2.contains(start) && l2.indexOf(start) > 0) start = l2.split("")[0];
}
}
ans += start;
// System.out.println(ans);
for (int i = 0; i < 10; i++) {
start = (getNext(slines, start));
if (start == null) break;
ans += start;
// System.out.println(ans);
}
Date endTime = new Date();
System.out.println(ans + " in " + (endTime.getTime() - startTime.getTime()) + " ms.");
}
/**
* Converts the form field values in the <tt>ffValuesIter</tt> into a caps string.
*
* @param ffValuesIter the {@link Iterator} containing the form field values.
* @param capsBldr a <tt>StringBuilder</tt> to which the caps string representing the form field
* values is to be appended
*/
private static void formFieldValuesToCaps(Iterator<String> ffValuesIter, StringBuilder capsBldr) {
SortedSet<String> fvs = new TreeSet<String>();
while (ffValuesIter.hasNext()) fvs.add(ffValuesIter.next());
for (String fv : fvs) capsBldr.append(fv).append('<');
}
/**
* Gets the unique read groups in the table
*
* @param report the GATKReport containing the table with RecalUtils.READGROUP_REPORT_TABLE_TITLE
* @return the unique read groups
*/
private static SortedSet<String> getReadGroups(final GATKReport report) {
final GATKReportTable reportTable = report.getTable(RecalUtils.READGROUP_REPORT_TABLE_TITLE);
final SortedSet<String> readGroups = new TreeSet<String>();
for (int i = 0; i < reportTable.getNumRows(); i++)
readGroups.add(reportTable.get(i, RecalUtils.READGROUP_COLUMN_NAME).toString());
return readGroups;
}
/**
* Returns all supported sample rates.
*
* @return an array of sample rates, in Hertz, never <code>null</code>.
*/
public Integer[] getSampleRates() {
final String rawValue = this.properties.get(DEVICE_SAMPLERATES);
final String[] values = rawValue.split(",\\s*");
final SortedSet<Integer> result =
new TreeSet<Integer>(
NumberUtils.<Integer>createNumberComparator(false /* aSortAscending */));
for (String value : values) {
result.add(Integer.valueOf(value.trim()));
}
return result.toArray(new Integer[result.size()]);
}
public static File[] getDirectoryListing(File dir) {
SortedSet<File> dirSet = new TreeSet<File>();
SortedSet<File> fileSet = new TreeSet<File>();
File[] files = dir.listFiles();
for (int i = 0; i < files.length; i++) {
if (files[i].isDirectory()) dirSet.add(files[i]);
else fileSet.add(files[i]);
}
List<File> fileList = new LinkedList<File>();
fileList.addAll(dirSet);
fileList.addAll(fileSet);
return fileList.toArray(new File[] {});
}
/**
* Prints one Unicode property value per line, along with its aliases, if any, for the given
* unicodeVersion.
*
* @param unicodeVersion The Unicode version to print property values and aliases for
* @throws UnicodeProperties.UnsupportedUnicodeVersionException if unicodeVersion is not supported
*/
private static void printUnicodePropertyValuesAndAliases(String unicodeVersion)
throws UnicodeProperties.UnsupportedUnicodeVersionException {
Pattern versionPattern = Pattern.compile("(\\d+)(?:\\.(\\d+))?(?:\\.\\d+)?");
Matcher matcher = versionPattern.matcher(unicodeVersion);
if (!matcher.matches()) {
throw new UnicodeProperties.UnsupportedUnicodeVersionException();
}
String underscoreVersion =
matcher.group(1) + (null == matcher.group(2) ? "_0" : "_" + matcher.group(2));
String[] propertyValues;
String[] propertyValueAliases;
try {
Class<?> clazz = Class.forName("jflex.unicode.data.Unicode_" + underscoreVersion);
Field field = clazz.getField("propertyValues");
propertyValues = (String[]) field.get(null);
field = clazz.getField("propertyValueAliases");
propertyValueAliases = (String[]) field.get(null);
} catch (Exception e) {
throw new UnicodeProperties.UnsupportedUnicodeVersionException();
}
SortedMap<String, SortedSet<String>> propertyValuesToAliases =
new TreeMap<String, SortedSet<String>>();
for (String value : propertyValues) {
propertyValuesToAliases.put(value, new TreeSet<String>());
}
for (int i = 0; i < propertyValueAliases.length; i += 2) {
String alias = propertyValueAliases[i];
String value = propertyValueAliases[i + 1];
SortedSet<String> aliases = propertyValuesToAliases.get(value);
if (null == aliases) {
aliases = new TreeSet<String>();
propertyValuesToAliases.put(value, aliases);
}
aliases.add(alias);
}
for (Map.Entry<String, SortedSet<String>> entry : propertyValuesToAliases.entrySet()) {
String value = entry.getKey();
SortedSet<String> aliases = entry.getValue();
Out.print(value);
if (aliases.size() > 0) {
for (String alias : aliases) {
Out.print(", " + alias);
}
}
Out.println("");
}
}
void clear() {
pending_entries.clear();
output_set.clear();
object_tasks.clear();
active_threads.clear();
time_marks.clear();
thread_entries = null;
next_time = 0;
end_time = 0;
current_thread = null;
thread_map.clear();
cpu_time = null;
thread_counter = 0;
task_counter = 0;
max_delta = 1;
}
public static String getNext(SortedSet<String> slines, String num) {
String nxt = "";
SortedSet<String> candidates = new TreeSet<String>();
for (String s : slines) {
if (s.contains(num) && s.indexOf(num) < s.length() - 1) {
candidates.add(s.split("")[s.indexOf(num) + 1]);
}
}
// System.out.println(candidates);
if (candidates.size() == 0) return null;
nxt = candidates.first();
for (String s : slines) {
for (String c : candidates) {
if (s.contains(c) && s.contains(nxt) && s.indexOf(c) < s.indexOf(nxt)) nxt = c;
}
}
return nxt;
}
/** Create a new set of shared job state counters for all jobs in the given group. */
public void initCounters(TaskTimer timer, QueueJobGroup group) {
long groupID = group.getGroupID();
if (LogMgr.getInstance().isLoggable(LogMgr.Kind.Ops, LogMgr.Level.Finest))
LogMgr.getInstance()
.log(LogMgr.Kind.Ops, LogMgr.Level.Finest, "Init Job Counts for Group (" + groupID + ")");
SortedSet<Long> jobIDs = group.getJobIDs();
Counters counters = new Counters(jobIDs.size());
timer.acquire();
synchronized (pCountersByGroup) {
timer.resume();
if (pCountersByGroup.put(groupID, counters) != null)
LogMgr.getInstance()
.logAndFlush(
LogMgr.Kind.Ops,
LogMgr.Level.Warning,
"Somehow the job group ("
+ groupID
+ ") was already in the state "
+ "counts table!");
}
timer.acquire();
synchronized (pCountersByJob) {
timer.resume();
for (Long jobID : jobIDs) {
if (pCountersByJob.put(jobID, counters) != null)
LogMgr.getInstance()
.logAndFlush(
LogMgr.Kind.Ops,
LogMgr.Level.Warning,
"Somehow the job (" + jobID + ") was already in the state counts table!");
}
}
}
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);
}
private BdynRangeSet addToRange(long start, long t0, long t1, BdynRangeSet rslt) {
OutputEntry timee = new OutputEntry(start);
SortedSet<OutputEntry> ss = output_set.tailSet(timee);
for (OutputEntry e1 : ss) {
if (e1.getStartTime() > t1) break;
if (e1.getEndTime(t1) >= t0) {
ThreadData td = e1.getThread();
if (rslt == null) rslt = new BdynRangeSet();
Set<BdynEntry> r1 = rslt.get(td);
if (r1 == null) {
r1 = new HashSet<BdynEntry>();
rslt.put(td, r1);
}
r1.add(e1);
}
}
return rslt;
}
/** disassembles program held in code store */
void disassembleProgram(String asmFileName) {
try {
asmOut = new FileWriter(asmFileName);
} catch (IOException e) {
System.out.println("Disassembler: can not create asm output file " + asmName);
error = true;
return;
}
// collect all addresses that may be the target of a jump instruction
SortedSet<Integer> targets = new TreeSet<Integer>();
for (int addr = Machine.CB; addr < Machine.CT; addr++) {
Instruction inst = Machine.code[addr];
Machine.Op op = Machine.intToOp[inst.op];
switch (op) {
case CALL:
// only consider calls (branches) within code memory (i.e. not primitives)
if (inst.r == Machine.Reg.CB.ordinal()) targets.add(inst.d);
break;
case JUMP:
// address following an unconditional branch is an implicit target
targets.add(addr + 1);
/* FALL THROUGH! */
case JUMPIF:
// a jump of any sort creates a branch target
targets.add(inst.d);
}
}
// map branch target addresses to unique labels
addrToLabel = new HashMap<Integer, String>();
int labelCounter = 10;
for (Integer addr : targets) {
String label = "L" + labelCounter++;
addrToLabel.put(addr, label);
}
// disassemble each instruction
for (int addr = Machine.CB; addr < Machine.CT; addr++) {
// generate instruction address
asmWrite(String.format("%3d ", addr));
// if this addr is a branch target, output label
if (addrToLabel.containsKey(addr))
asmWrite(String.format("%-7s", addrToLabel.get(addr) + ":"));
else asmWrite(" ");
// instruction
writeInstruction(Machine.code[addr]);
// newline
asmWrite("\n");
}
// close output file
try {
asmOut.close();
} catch (IOException e) {
error = true;
}
}
Iterator<Long> getTimeMarkIterator() {
return time_marks.iterator();
}
void addTimeMark(long when) {
time_marks.add(when);
}
int getActiveThreadCount() {
return active_threads.size();
}
long getStartTime() {
if (output_set.isEmpty()) return 0;
return output_set.first().getStartTime();
}
public static QueryFilter namesQueryFilter(
ColumnFamilyStore cfs, DecoratedKey key, CellName... names) {
SortedSet<CellName> s = new TreeSet<CellName>(cfs.getComparator());
for (CellName n : names) s.add(n);
return QueryFilter.getNamesFilter(key, cfs.name, s, System.currentTimeMillis());
}
public static NamesQueryFilter namesFilter(ColumnFamilyStore cfs, String... names) {
SortedSet<CellName> s = new TreeSet<CellName>(cfs.getComparator());
for (String str : names) s.add(cellname(str));
return new NamesQueryFilter(s);
}
/**
* Calculates the <tt>String</tt> for a specific <tt>DiscoverInfo</tt> which is to be hashed in
* order to compute the ver string for that <tt>DiscoverInfo</tt>.
*
* @param discoverInfo the <tt>DiscoverInfo</tt> for which the <tt>String</tt> to be hashed in
* order to compute its ver string is to be calculated
* @return the <tt>String</tt> for <tt>discoverInfo</tt> which is to be hashed in order to compute
* its ver string
*/
private static String calculateEntityCapsString(DiscoverInfo discoverInfo) {
StringBuilder bldr = new StringBuilder();
// Add identities
{
Iterator<DiscoverInfo.Identity> identities = discoverInfo.getIdentities();
SortedSet<DiscoverInfo.Identity> is =
new TreeSet<DiscoverInfo.Identity>(
new Comparator<DiscoverInfo.Identity>() {
public int compare(DiscoverInfo.Identity i1, DiscoverInfo.Identity i2) {
int category = i1.getCategory().compareTo(i2.getCategory());
if (category != 0) return category;
int type = i1.getType().compareTo(i2.getType());
if (type != 0) return type;
/*
* TODO Sort by xml:lang.
*
* Since sort by xml:lang is currently missing,
* use the last supported sort criterion i.e.
* type.
*/
return type;
}
});
if (identities != null) while (identities.hasNext()) is.add(identities.next());
for (DiscoverInfo.Identity i : is) {
bldr.append(i.getCategory())
.append('/')
.append(i.getType())
.append("//")
.append(i.getName())
.append('<');
}
}
// Add features
{
Iterator<DiscoverInfo.Feature> features = getDiscoverInfoFeatures(discoverInfo);
SortedSet<String> fs = new TreeSet<String>();
if (features != null) while (features.hasNext()) fs.add(features.next().getVar());
for (String f : fs) bldr.append(f).append('<');
}
DataForm extendedInfo = (DataForm) discoverInfo.getExtension("x", "jabber:x:data");
if (extendedInfo != null) {
synchronized (extendedInfo) {
SortedSet<FormField> fs =
new TreeSet<FormField>(
new Comparator<FormField>() {
public int compare(FormField f1, FormField f2) {
return f1.getVariable().compareTo(f2.getVariable());
}
});
FormField formType = null;
for (Iterator<FormField> fieldsIter = extendedInfo.getFields(); fieldsIter.hasNext(); ) {
FormField f = fieldsIter.next();
if (!f.getVariable().equals("FORM_TYPE")) fs.add(f);
else formType = f;
}
// Add FORM_TYPE values
if (formType != null) formFieldValuesToCaps(formType.getValues(), bldr);
// Add the other values
for (FormField f : fs) {
bldr.append(f.getVariable()).append('<');
formFieldValuesToCaps(f.getValues(), bldr);
}
}
}
return bldr.toString();
}
public static ImmutableSortedSet copyOfSorted(SortedSet sortedset) {
Comparator comparator1 = sortedset.comparator();
if (comparator1 == null) comparator1 = NATURAL_ORDER;
return copyOfInternal(comparator1, sortedset);
}
/**
* Translates a StatValue into a PerfStatValue with the specified filter, with the specified
* operations, and the specified trim. For the latter, if start is -1, uses the beginning of the
* statistic's existence, and if end is -1, goes to the end of the statistic's existence.
*/
protected static PerfStatValue getPerfStatValue(
StatSpec statspec, TrimSpec trimspec, StatArchiveReader.StatValue sv) {
long start = trimspec.getStart();
long end = trimspec.getEnd();
sv = sv.createTrimmed(start, end);
if (Log.getLogWriter().finestEnabled()) {
Log.getLogWriter()
.finest(
"PerfStatReader: Trimmed from "
+ trimspec.getStartStr()
+ " ("
+ start
+ ") to "
+ trimspec.getEndStr()
+ " ("
+ end
+ ")");
}
int filter = statspec.getFilter();
double mean = -1;
if (filter == StatArchiveReader.StatValue.FILTER_PERSEC && statspec.getMean()) {
if (start == -1) {
start = sv.getRawAbsoluteTimeStamps()[0];
}
if (end == -1) {
long[] rats = sv.getRawAbsoluteTimeStamps();
end = rats[rats.length - 1];
}
long elapsedSec = (end - start) / 1000;
sv.setFilter(StatArchiveReader.StatValue.FILTER_NONE);
double del = sv.getSnapshotsMaximum() - sv.getSnapshotsMinimum();
mean = del / elapsedSec;
}
sv.setFilter(filter);
// @todo lises see if psv really needs to hang onto specs
PerfStatValue psv = new PerfStatValue(statspec, trimspec);
psv.setIsLargerBetter(sv.getDescriptor().isLargerBetter());
psv.setSamples(sv.getSnapshotsSize());
if (statspec.getMin()) psv.setMin(sv.getSnapshotsMinimum());
if (statspec.getMax()) psv.setMax(sv.getSnapshotsMaximum());
if (statspec.getMaxMinusMin())
psv.setMaxMinusMin(sv.getSnapshotsMaximum() - sv.getSnapshotsMinimum());
if (statspec.getMean()) {
if (filter == StatArchiveReader.StatValue.FILTER_PERSEC) {
psv.setMean(mean);
} else {
psv.setMean(sv.getSnapshotsAverage());
}
}
if (statspec.getStddev()) psv.setStddev(sv.getSnapshotsStandardDeviation());
SortedSet archives = new TreeSet();
StatArchiveReader.ResourceInst[] resources = sv.getResources();
String productVersion = null;
for (int i = 0; i < resources.length; i++) {
String archive = resources[i].getArchive().getFile().getParentFile().getName();
if (productVersion == null) {
productVersion = resources[i].getArchive().getArchiveInfo().getProductVersion();
}
if (!archives.contains(archive)) {
archives.add(archive);
}
}
psv.setArchives(archives);
psv.setProductVersion(productVersion);
return psv;
}
/**
* Collects the differentially expressed segments (segment pairs with the same start and end, but
* one member of the pair has the LINE type while the other does not) from a given segmentation.
*
* @param seg
* @param c1
* @param c2
* @return
*/
private Collection<DifferentialKey> differentialRegions(InputSegmentation seg, int c1, int c2) {
ArrayList<DifferentialKey> keys = new ArrayList<DifferentialKey>();
// Get the ground-truth data, so we can assemble the region identifiers when
// we've found the differentially expressed segments.
InputData data = seg.input;
Integer[] locations = data.locations();
String strand = data.strand();
String chrom = data.chrom();
// Most of the setup of this method is filtering and sorting the Segment objects,
// so that we *only* consider the segments from the two given channels (c1, and c2)
// and so that we consider them in order, so that we're not comparing segments
// different parts of the input chunk.
SortedSet<Segment> c1Segs = new TreeSet<Segment>();
SortedSet<Segment> c2Segs = new TreeSet<Segment>();
for (Segment s : seg.segments) {
if (s.channel == c1) {
c1Segs.add(s);
} else if (s.channel == c2) {
c2Segs.add(s);
}
}
// These should now be in sorted order, given that we built them using the
// SortedSet objects.
Segment[] c1array = c1Segs.toArray(new Segment[0]);
Segment[] c2array = c2Segs.toArray(new Segment[0]);
// sanity check.
if (c1array.length != c2array.length) {
throw new IllegalArgumentException();
}
for (int i = 0; i < c1array.length; i++) {
Segment s1 = c1array[i], s2 = c2array[i];
// sanity check.
if (!s1.start.equals(s2.start) || !s1.end.equals(s2.end)) {
throw new IllegalArgumentException(
String.format("%d,%d doesn't match %d,%d", s1.start, s1.end, s2.start, s2.end));
}
// There are three conditions here:
// (1) Neither of the segments is 'shared'
// (2) The segments don't have the same type
// (3) At least one of them is a line.
if (!s1.shared && !s2.shared) {
if (s1.segmentType.equals(Segment.LINE) || s2.segmentType.equals(Segment.LINE)) {
boolean differential = false;
// Remember: for any segment 's', s.start and s.end are *indices*
// into the 'locations' array of the corresponding data chunk.
DifferentialKey key =
new DifferentialKey(
new RegionKey(chrom, locations[s1.start], locations[s1.end], strand), s1, s2);
if (s1.segmentType.equals(Segment.FLAT)) {
differential = key.s2Expr() > key.s1Expr();
} else if (s2.segmentType.equals(Segment.FLAT)) {
differential = key.s1Expr() > key.s2Expr();
} else {
differential = Math.abs(key.diffExpr()) >= 0.5;
}
if (differential) {
// If all conditions have been satisfied, then we build
// the region identifier and save it in the list to be returned.
keys.add(key);
}
}
}
}
return keys;
}
