/**
* Given a graph database gDB, first prune distinct edges with the edgeIndex Then randomly
* selected chooseN graphs
*
* @param gDB
* @param chooseN
* @param pruneDistinctEdges
* @param edgeIndex
* @return
* @throws IOException
* @throws ParseException
*/
public static Graph[] randomlyChooseDBGraph(
GraphDatabase_OnDisk gDB, int chooseN, boolean pruneDistinctEdges, EdgeIndex edgeIndex)
throws IOException, ParseException {
if (pruneDistinctEdges == false) return randomlyChooseDBGraph(gDB, chooseN);
// First Step: filter out graphs that are not in between the boundary
String tempFileName = gDB.getDBFileName() + "_temp";
BufferedWriter tempDBWriter = new BufferedWriter(new FileWriter(tempFileName));
int count = 0;
float edgeNum = 0, nodeNum = 0;
for (int i = 0; i < gDB.getTotalNum(); i++) {
Graph theGraph = gDB.findGraph(i);
if (edgeIndex.containInfrequentEdges(theGraph)) continue;
else {
tempDBWriter.write(count++ + " => " + gDB.findGraphString(i) + "\n");
edgeNum += theGraph.getEdgeCount();
nodeNum += theGraph.getNodeCount();
}
}
tempDBWriter.close();
// Intrigue java garbage collector
Runtime r = Runtime.getRuntime();
r.gc();
// Write the meta information of the new file
BufferedWriter metaWriter = new BufferedWriter(new FileWriter(tempFileName + "_Meta"));
// 1. Processing Date
SimpleDateFormat bartDateFormat = new SimpleDateFormat("EEEE-MMMM-dd-yyyy");
Date date = new Date();
metaWriter.write(bartDateFormat.format(date));
metaWriter.newLine();
// 2. Number of graphs in this file
metaWriter.write("Number of Graphs:" + count);
metaWriter.newLine();
metaWriter.write(
"Average EdgeNum: "
+ (float) (edgeNum) / count
+ ", Average NodeNum: "
+ (float) (nodeNum) / count);
// Close meta data file
try {
metaWriter.close();
} catch (IOException e) {
e.printStackTrace();
}
return randomlyChooseDBGraph(new GraphDatabase_OnDisk(tempFileName, gDB.getParser()), chooseN);
}
@Override
public Document vectorize(int gID, Graph g, Map<Integer, String> gHash) throws ParseException {
Document gDoc = new Document();
if (g.getEdgeCount() == 0) return gDoc;
String graphString = gParser.serialize(g);
Field stringField = new Field("gString", graphString, Field.Store.YES, Field.Index.NO);
gDoc.add(stringField);
Field IDField = new Field("gID", new Integer(gID).toString(), Field.Store.YES, Field.Index.NO);
gDoc.add(IDField);
List<Integer> allIDs = searcher.subgraphs(g, new SearchStatus());
// 0. Add One "-1" to the subGraphs fields [for pure mustNot search]
gDoc.add(
new Field(
"subGraphs", (new Integer(-1)).toString(), Field.Store.NO, Field.Index.NOT_ANALYZED));
if (allIDs == null || allIDs.size() == 0) return gDoc;
// 1.
Collections.sort(allIDs);
for (int i = 0; i < allIDs.size(); i++) {
if (gHash == null) {
String byteString = allIDs.get(i).toString();
gDoc.add(new Field("subGraphs", byteString, Field.Store.NO, Field.Index.NOT_ANALYZED));
} else {
gDoc.add(
new Field(
"subGraphs", gHash.get(allIDs.get(i)), Field.Store.NO, Field.Index.NOT_ANALYZED));
}
}
// StringBuffer sBuf = new StringBuffer();
// for(int i = 0; i< allIDs.length; i++ ){
// if(gHash==null)
// sBuf.append(new Integer(allIDs[i]).toString());
// else {
// sBuf.append(gHash.get(allIDs[i]));
// // //TEST
// // String indexString = gHash.get(allIDs[i]);
// // String indexString2 = searcher.getLabel(allIDs[i]);
// // if(!indexString.equals(indexString2)){
// // System.out.println("lala");
// // }
// // //END OF TEST
// }
// sBuf.append(" ");
// }
// String termString = sBuf.substring(0, sBuf.length()-1);
//
// Field subgraphField = new Field("subGraphs", termString,
// Field.Store.NO, Field.Index.ANALYZED_NO_NORMS);
// gDoc.add(subgraphField);
return gDoc;
}
public static void saveGDB(Graph[] graphs, GraphParser gParser, String fileName)
throws IOException {
// First Step: filter out graphs that are not in between the boundary
BufferedWriter dbWriter = new BufferedWriter(new FileWriter(fileName));
int count = 0;
float edgeNum = 0, nodeNum = 0;
for (int i = 0; i < graphs.length; i++) {
Graph theGraph = graphs[i];
dbWriter.write(count++ + " => " + gParser.serialize(theGraph) + "\n");
edgeNum += theGraph.getEdgeCount();
nodeNum += theGraph.getNodeCount();
}
dbWriter.close();
// Intrigue java garbage collector
Runtime r = Runtime.getRuntime();
r.gc();
// Write the meta information of the new file
BufferedWriter metaWriter = new BufferedWriter(new FileWriter(fileName + "_Meta"));
// 1. Processing Date
SimpleDateFormat bartDateFormat = new SimpleDateFormat("EEEE-MMMM-dd-yyyy");
Date date = new Date();
metaWriter.write(bartDateFormat.format(date));
metaWriter.newLine();
// 2. Number of graphs in this file
metaWriter.write("Number of Graphs:" + count);
metaWriter.newLine();
metaWriter.write(
"Average EdgeNum: "
+ (float) (edgeNum) / count
+ ", Average NodeNum: "
+ (float) (nodeNum) / count);
// Close meta data file
try {
metaWriter.close();
} catch (IOException e) {
e.printStackTrace();
}
}
public static void changeFormatSimple(
String SDFFileName, String SmilesFileName, GraphFactory graphFactory) throws IOException {
// Open InputFile
BufferedReader fileBufReader = new BufferedReader(new FileReader(SDFFileName));
// Open OutputFile
BufferedWriter outputWriter = new BufferedWriter(new FileWriter(SmilesFileName));
SDFParserModified SDFParser = MyFactory.getSDFParserM();
SmilesParser smilesParser = MyFactory.getSmilesParser();
Graph oneGraph = null;
double aveEdgeCount = 0;
double aveNodeCount = 0;
int index = 0;
String spliter = " => ";
while (fileBufReader.ready()) {
try {
oneGraph = SDFParser.parse(fileBufReader, graphFactory);
} catch (ParseException e) {
System.out.println("skip one graph");
while (fileBufReader.ready()) {
String aLine = fileBufReader.readLine();
if (aLine.equals("$$$$")) break;
}
continue;
}
if (GraphConnectivityTester.isConnected(oneGraph)) {
aveEdgeCount =
(aveEdgeCount) * (index / ((double) index + 1))
+ oneGraph.getEdgeCount() / ((double) index + 1);
aveNodeCount =
(aveNodeCount) * (index / ((double) index + 1))
+ oneGraph.getNodeCount() / ((double) index + 1);
if (index > 0) outputWriter.newLine();
outputWriter.write(index + spliter + smilesParser.serialize(oneGraph));
index++;
}
}
// Close input File
try {
fileBufReader.close();
outputWriter.flush();
outputWriter.close();
} catch (IOException e) {
e.printStackTrace();
}
System.out.println(
"In processor: changeFormat, "
+ index
+ " number of graphs"
+ "has been formated into Smiles Format");
// Intrigue java garbage collector
Runtime r = Runtime.getRuntime();
r.gc();
// Write the meta information of the smile data file:
BufferedWriter metaWriter = new BufferedWriter(new FileWriter(SmilesFileName + "_Meta"));
// 1. Processing Date
SimpleDateFormat bartDateFormat = new SimpleDateFormat("EEEE-MMMM-dd-yyyy");
Date date = new Date();
metaWriter.write(bartDateFormat.format(date));
metaWriter.newLine();
// 2. Number of graphs in this file
metaWriter.write("Number of Graphs:" + index);
metaWriter.write("Ave EdgeCount: " + aveEdgeCount + " Ave NodeCount: " + aveNodeCount);
// Close meta data file
try {
metaWriter.flush();
metaWriter.close();
} catch (IOException e) {
e.printStackTrace();
}
}
/**
* @param smilesDBFile
* @param chooseN
* @param chosenDBFile
* @throws ParseException
* @throws IOException
*/
public static void randomlyChooseTwoDBGraph(String smilesDBFile, int chooseN, String chosenDBFile)
throws ParseException, IOException {
// First get the number of graphs in this database, which can be found in the metadata file
BufferedReader metaFile = new BufferedReader(new FileReader(smilesDBFile + "_Meta"));
metaFile.readLine();
String[] temp = metaFile.readLine().split(":");
int m = Integer.parseInt(temp[1]);
metaFile.close();
int edgeNum = 0, nodeNum = 0;
// Then randomly select n from m: the first chooseN indexes are sorted as stored in indexes
int[] indexes = new int[m];
for (int i = 0; i < m; i++) indexes[i] = i;
Random rd = new Random();
int j = 0;
int swapTemp = 0;
;
for (int i = 0; i < 2 * chooseN; i++) {
j = (int) (rd.nextFloat() * (m - i)) + i;
swapTemp = indexes[i];
indexes[i] = indexes[j];
indexes[j] = swapTemp;
}
Arrays.sort(indexes, 0, 2 * chooseN);
// Read those db graphs and save them into the new file
BufferedReader fullDBReader = new BufferedReader(new FileReader(smilesDBFile));
BufferedWriter chosenDBWriter1 = new BufferedWriter(new FileWriter(chosenDBFile + "_1"));
BufferedWriter chosenDBWriter2 = new BufferedWriter(new FileWriter(chosenDBFile + "_2"));
int fileLineIndex = 0;
int i = 0;
String aLine = null;
String spliter = " => ";
while ((aLine = fullDBReader.readLine()) != null && i < 2 * chooseN) {
if (fileLineIndex < indexes[i]) {
fileLineIndex++;
continue; // keep on reading
} else if (fileLineIndex == indexes[i] && i < chooseN) {
// index=> orignalIndex =>smiles
String gString = aLine.split(spliter)[1];
Graph g = MyFactory.getSmilesParser().parse(gString, MyFactory.getGraphFactory());
edgeNum += g.getEdgeCount();
nodeNum += g.getNodeCount();
chosenDBWriter1.write(i + spliter + gString);
chosenDBWriter1.newLine();
i++;
fileLineIndex++;
} else if (fileLineIndex == indexes[i] && i >= chooseN) {
// index=> orignalIndex =>smiles
String gString = aLine.split(spliter)[1];
Graph g = MyFactory.getSmilesParser().parse(gString, MyFactory.getGraphFactory());
edgeNum += g.getEdgeCount();
nodeNum += g.getNodeCount();
chosenDBWriter2.write(i + spliter + gString);
chosenDBWriter2.newLine();
i++;
fileLineIndex++;
} else if (fileLineIndex > indexes[i])
System.out.println("Exception: Processor: randomlyChooseDBGraph");
}
// Close out File
try {
chosenDBWriter1.close();
chosenDBWriter2.close();
fullDBReader.close();
} catch (IOException e) {
e.printStackTrace();
}
// Intrigue java garbage collector
Runtime r = Runtime.getRuntime();
r.gc();
// Write the meta information of the new file
BufferedWriter metaWriter = new BufferedWriter(new FileWriter(chosenDBFile + "_1" + "_Meta"));
// 1. Processing Date
SimpleDateFormat bartDateFormat = new SimpleDateFormat("EEEE-MMMM-dd-yyyy");
Date date = new Date();
metaWriter.write(bartDateFormat.format(date));
metaWriter.newLine();
// 2. Number of graphs in this file
metaWriter.write("Number of Graphs:" + i);
metaWriter.newLine();
metaWriter.write(
"Average EdgeNum: "
+ (float) (edgeNum) / i
+ ", Average NodeNum: "
+ (float) (nodeNum) / i);
// Close meta data file
try {
metaWriter.close();
} catch (IOException e) {
e.printStackTrace();
}
}