2
|
1 package edu.unc.genomics.ngs;
|
|
2
|
|
3 import java.io.BufferedWriter;
|
|
4 import java.io.IOException;
|
|
5 import java.nio.charset.Charset;
|
|
6 import java.nio.file.Files;
|
|
7 import java.nio.file.Path;
|
|
8
|
|
9 import org.apache.commons.math.stat.Frequency;
|
|
10 import org.apache.log4j.Logger;
|
|
11
|
|
12 import com.beust.jcommander.Parameter;
|
|
13
|
|
14 import edu.unc.genomics.CommandLineTool;
|
|
15 import edu.unc.genomics.Interval;
|
|
16 import edu.unc.genomics.io.IntervalFile;
|
|
17
|
|
18 public class IntervalLengthDistribution extends CommandLineTool {
|
|
19
|
|
20 private static final Logger log = Logger.getLogger(IntervalLengthDistribution.class);
|
|
21
|
|
22 @Parameter(names = {"-i", "--input"}, description = "Interval file", required = true)
|
|
23 public IntervalFile<? extends Interval> inputFile;
|
|
24 @Parameter(names = {"-o", "--output"}, description = "Output file", required = true)
|
|
25 public Path outputFile;
|
|
26
|
|
27
|
|
28 @Override
|
|
29 public void run() throws IOException {
|
|
30 log.debug("Generating histogram of interval lengths");
|
|
31 Frequency freq = new Frequency();
|
|
32 int min = Integer.MAX_VALUE;
|
|
33 int max = -1;
|
|
34 for (Interval i : inputFile) {
|
|
35 int L = i.length();
|
|
36 freq.addValue(L);
|
|
37
|
|
38 if (L < min) {
|
|
39 min = L;
|
|
40 }
|
|
41 if (L > max) {
|
|
42 max = L;
|
|
43 }
|
|
44 }
|
|
45
|
|
46 log.debug("Writing histogram output");
|
|
47 try (BufferedWriter writer = Files.newBufferedWriter(outputFile, Charset.defaultCharset())) {
|
|
48 for (int i = min; i <= max; i++) {
|
|
49 writer.write(i+"\t"+freq.getCount(i));
|
|
50 writer.newLine();
|
|
51 }
|
|
52 }
|
|
53 }
|
|
54
|
|
55 public static void main(String[] args) {
|
|
56 new IntervalLengthDistribution().instanceMain(args);
|
|
57 }
|
|
58
|
|
59 }
|