层次聚类算法的java实现

package cluster; import java.util.*; import java.util.regex.Pattern; import java.io.*; public class MyClusterByClique { private static int cellRow; private static int cellColumn; private static int numberClusters; private static Cell[][] cells; private static List clusters; public static void main(String[] args) { try { initialize(); read(); cluster(); printResult(); System.out.println("Cluster by Clique is over.:)"); } catch (Exception e) { System.err.println("Err in main() ::" + e); } } private static void initialize() { try { cellRow = ClusterConfig.cellRow; cellColumn = ClusterConfig.cellColumn; numberClusters = 0; cells = new Cell[cellRow + 1][cellColumn + 1]; clusters = new ArrayList(); clusters.add(new Cluster(numberClusters)); int i, j; for (i = 1; i <= ClusterConfig.cellRow; i++) { for (j = 1; j <= ClusterConfig.cellColumn; j++) { cells[i][j] = new Cell(); } } } catch (Exception e) { System.err.println("Err in ::" + e); } }// initialize() end; private static void read() { try { BufferedReader inByLine = new BufferedReader(new FileReader( ClusterConfig.filePath)); System.out.println("打开文件成功: " + ClusterConfig.filePath + " opened"); System.out.println("开始读取: Read() begin"); String piontStr = null; int i = 0; while ((piontStr = inByLine.readLine()) != null) { double x; double y; int cellx; int celly; String[] pointsStr = Pattern.compile("\\s").split(piontStr); i++; x = Double.valueOf(pointsStr[0]); y = Double.valueOf(pointsStr[1]); cellx = (int) Math.ceil((x - ClusterConfig.xmin) / ClusterConfig.intervalx); celly = (int) Math.ceil((y - ClusterConfig.ymin) / ClusterConfig.intervaly); cells[celly][cellx].addnumberPoints(); if (cells[celly][cellx].getNumberPoints() >= ClusterConfig.densityThreshold) cells[celly][cellx].setQualified(1); System.out.println("读入点 " + i + "# : " + x + " , " + y); } inByLine.close(); System.out.println("读取结束:Read() finished "); } catch (Exception e) { System.err.println("Err in read() ::" + e); } }// read() end private static void cluster() { try { int i, j; for (i = 1; i <= ClusterConfig.cellRow; i++) { for (j = 1; j <= ClusterConfig.cellColumn; j++) { if (cells[i][j].getClusterNo() == 0 && cells[i][j].getChecked() != 1) { if (cells[i][j].getQualified() > 0) { numberClusters += 1; clusters.add(new Cluster(numberClusters)); retrieve(numberClusters,i,j); System.out.println("Cluster():i = " + i + " , j = " + j+"ClusterNo = "+ ((Cluster)clusters.get(numberClusters)).getClusterNo() ); } else { cells[i][j].setChecked(1); cells[i][j].setClusterNo(0); }// end if }// end if }// end for }// end for } catch (Exception e) { System.err.println("Err in cluster() ::" + e); } }// cluster() end; private static void retrieve(int k,int i ,int j) { try { int l, m; if (cells[i][j].getChecked() != 1 && cells[i][j].getClusterNo() == 0) { System.out.println("infor in retireve( i ="+ i + ", j ="+ j+ " ) :: nubmer of points = "+cells[i][j].getNumberPoints()+" , number of threshold = "+(int) ((Cluster)clusters.get(k)).getDensity()* ClusterConfig.densityRatioThreshold); if (cells[i][j].getQualified() > 0 && cells[i][j].getNumberPoints() >= (int) ((Cluster) clusters.get(k)).getDensity()* ClusterConfig.densityRatioThreshold) { cells[i][j].setChecked(1); cells[i][j].setClusterNo(k); ((Cluster) clusters.get(k)).addCell(cells[i][j] .getNumberPoints()); // cluster the neighbors of cells[i][j] if (i != 1) { l = i - 1; retrieve(k, l, j); } if (i != cellColumn) { l = i + 1; retrieve(k,l,j); } if (j != 1) { m = j - 1; retrieve(k,i, m); } if (j != cellRow) { m = j + 1; retrieve(k, i, m); } } else { cells[i][j].setChecked(1); cells[i][j].setClusterNo(0); } } } catch (Exception e) { System.err.println("Err in retrieve() ::" + e); } }// retrieve() end; private static void printResult() { try { PrintWriter out = new PrintWriter(new BufferedWriter( new FileWriter(ClusterConfig.outPath))); out.println("*******************aurthor Email.angus006@gmail.com××××××××××××××××××××××××"); for (int j = 0; j <= cellColumn; j++) out.printf("%7d", j); out.println("\n"); for (int i = 1; i <= cellRow; i++) { out.printf("%7d", i); for (int j = 1; j <= cellColumn; j++){ //out.printf("%2d", gv.cells[i][j].getClusterNo()); //out.printf(":"); out.printf("%7d",cells[i][j].getNumberPoints()); } out.println("\n"); } System.out.println("输出单元点数信息完毕，结果写入文件" + ClusterConfig.outPath); out.println("\n"); out.println("\n"); out.println("\n"); out.println("\n"); out.println("\n"); out.println("\n"); for (int j = 0; j <= cellColumn; j++) out.printf("%7d", j); out.println("\n"); for (int i = 1; i <= cellRow; i++) { out.printf("%7d", i); for (int j = 1; j <= cellColumn; j++){ out.printf("%2d", cells[i][j].getClusterNo()); out.printf("%7d",cells[i][j].getNumberPoints()); out.printf(":"); } out.println("\n"); } out.println("下面是各点详细信息，其中类别 0 表示离散点"); BufferedReader inByLine = new BufferedReader(new FileReader( ClusterConfig.filePath)); System.out.println("打开文件成功: " + ClusterConfig.filePath + " opened"); System.out.println("开始读取: Read() begin"); String piontStr = null; int i = 0; while ((piontStr = inByLine.readLine()) != null) { double x; double y; int cellx; int celly; String[] pointsStr = Pattern.compile("\\s").split(piontStr); i++; x = Double.valueOf(pointsStr[0]); y = Double.valueOf(pointsStr[1]); cellx = (int) Math.ceil((x - ClusterConfig.xmin) / ClusterConfig.intervalx); celly = (int) Math.ceil((y - ClusterConfig.ymin) / ClusterConfig.intervaly); out.println("点 ( "+x+" , "+y + " ) , 类别：" +cells.clone()[celly][cellx].getClusterNo()+"; 该点位于 行号为 "+celly+ " ,列号为 "+cellx+"的单元格内."); } System.out.println("输出单元聚类信息完毕，结果写入文件" + ClusterConfig.outPath); inByLine.close(); out.close(); } catch (Exception e) { System.err.println("Err in printResult() ::" + e); } }// printResult() end; }