/src/edu/psu/chemxseer/structure/preprocess/RandomChoseDBGraph.java

package edu.psu.chemxseer.structure.preprocess;

import java.io.BufferedReader;
import java.io.BufferedWriter;
import java.io.FileReader;
import java.io.FileWriter;
import java.io.IOException;
import java.text.ParseException;
import java.text.SimpleDateFormat;
import java.util.Arrays;
import java.util.Date;
import java.util.Random;

import de.parmol.graph.Graph;
import de.parmol.parsers.GraphParser;
import edu.psu.chemxseer.structure.subsearch.FGindex.EdgeIndex;
import edu.psu.chemxseer.structure.subsearch.Impl.GraphDatabase_OnDisk;
import edu.psu.chemxseer.structure.subsearch.Interfaces.GraphDatabase;

public class RandomChoseDBGraph {

	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();
		}
	}
	/**
	 * 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);

	}

	
	
	/**
	 * Given a graph database gDB, randomly chooseN graphs, then return those graphs
	 * @param gDB
	 * @param chooseN
	 * @return
	 * @throws ParseException
	 * @throws IOException
	 */
	public static Graph[] randomlyChooseDBGraph(GraphDatabase gDB, int chooseN) throws ParseException, IOException{	
		// First get the number of graphs in this database, which can be found in the metadata file
		int m = gDB.getTotalNum();
		//System.out.println("Choose " + chooseN + " from " + m);
		if(m < chooseN){
			System.out.println("There is not need to choseN, the database is smaller than chooseN");
			return null;
		}
		// 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< chooseN; i++){
			j = (int) (rd.nextFloat() * (m-i))+i;
			swapTemp = indexes[i];
			indexes[i]=indexes[j];
			indexes[j] = swapTemp;
		}
		Arrays.sort(indexes, 0, chooseN);
		// Read those db graphs and save them into the new file
		Graph[] results = new Graph[chooseN];
		for(int i = 0; i< chooseN; i++){
			int gID = indexes[i];
			results[i] = gDB.findGraph(gID);
		}
		
		// Intrigue java garbage collector
		Runtime r = Runtime.getRuntime();
		r.gc();
		return results;
	}
	
	/**
	 * Given a graph database gDB, randomly choose N graphs, and then store them in the chosen database
	 * @param gDB
	 * @param chooseN
	 * @param chosenDBName
	 * @return
	 * @throws IOException 
	 */
	public static void randomlyChooseDBGraph(GraphDatabase gDB, int chooseN, String chosenDBName) throws IOException{
		// First get the number of graphs in this database, which can be found in the metadata file
		int m = gDB.getTotalNum();
		System.out.println("Choose " + chooseN + " from " + m);
		if(m < chooseN){
			System.out.println("There is not need to choseN, the database is smaller than chooseN");
			return;
		}
		// 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< chooseN; i++){
			j = (int) (rd.nextFloat() * (m-i))+i;
			swapTemp = indexes[i];
			indexes[i]=indexes[j];
			indexes[j] = swapTemp;
		}
		Arrays.sort(indexes, 0, chooseN);
		BufferedWriter chosenDBWriter = new BufferedWriter(new FileWriter(chosenDBName));
		String spliter = " => ";
		// Read those db graphs and save them into the new file
		for(int i = 0; i< chooseN; i++){
			int gID = indexes[i];
			String gString = gDB.findGraphString(gID);
			chosenDBWriter.write(i + spliter +gString);
			chosenDBWriter.newLine();
		}
		chosenDBWriter.close();
		// Intrigue java garbage collector
		Runtime r = Runtime.getRuntime();
		r.gc();
		
		BufferedWriter metaWriter = new BufferedWriter(new FileWriter(chosenDBName+ "_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:" + chooseN);
		metaWriter.newLine();
		metaWriter.write("Average EdgeNum: " + 0 + ", Average NodeNum: " + 
				0);
		// Close meta data file
		try {
			metaWriter.close();
		} catch (IOException e) {
			e.printStackTrace();
		}
	}
	
	
	public static void randomlySplitDBGraph(GraphDatabase gDB, int chooseN, String chosenDBName, 
			String leftDBName) throws IOException{
		// First get the number of graphs in this database, which can be found in the metadata file
		int m = gDB.getTotalNum();
		System.out.println("Choose " + chooseN + " from " + m);
		if(m < chooseN){
			System.out.println("There is not need to choseN, the database is smaller than chooseN");
			return;
		}
		// 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< chooseN; i++){
			j = (int) (rd.nextFloat() * (m-i))+i;
			swapTemp = indexes[i];
			indexes[i]=indexes[j];
			indexes[j] = swapTemp;
		}
		Arrays.sort(indexes, 0, chooseN);
		Arrays.sort(indexes, chooseN, m);
		BufferedWriter chosenDBWriter = new BufferedWriter(new FileWriter(chosenDBName));
		String spliter = " => ";
		// Read those db graphs and save them into the new file
		for(int i = 0; i< chooseN; i++){
			int gID = indexes[i];
			String gString = gDB.findGraphString(gID);
			chosenDBWriter.write(i + spliter +gString);
			chosenDBWriter.newLine();
		}
		chosenDBWriter.close();
		// Read those db graphs and save them into the new file
		BufferedWriter leftDBWriter = new BufferedWriter(new FileWriter(leftDBName));
		for(int i = chooseN; i< m; i++){
			int gID = indexes[i];
			String gString = gDB.findGraphString(gID);
			leftDBWriter.write((i-chooseN) + spliter +gString);
			leftDBWriter.newLine();
		}
		leftDBWriter.close();
		// Intrigue java garbage collector
		Runtime r = Runtime.getRuntime();
		r.gc();
		
		BufferedWriter metaWriter = new BufferedWriter(new FileWriter(chosenDBName+ "_Meta"));
		BufferedWriter metaWriter2 = new BufferedWriter(new FileWriter(leftDBName + "_Meta"));
		// 1. Processing Date
		SimpleDateFormat bartDateFormat = new SimpleDateFormat("EEEE-MMMM-dd-yyyy"); 
		Date date = new Date(); 
		metaWriter.write(bartDateFormat.format(date));
		metaWriter.newLine();
		metaWriter2.write(bartDateFormat.format(date));
		metaWriter2.newLine();
		// 2. Number of graphs in this file
		metaWriter.write("Number of Graphs:" + chooseN);
		metaWriter.newLine();
		metaWriter.write("Average EdgeNum: " + 0 + ", Average NodeNum: " + 
				0);
		metaWriter2.write("Number of Graphs:" + (m-chooseN));
		metaWriter2.newLine();
		metaWriter2.write("Average EdgeNum: " + 0 + ", Average NodeNum: " + 
				0);
		// Close meta data file
		try {
			metaWriter.close();
			metaWriter2.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();
		}
	}

}