Refine mean/median sequence length output

test/integration/FileContentsTest_complex_fastqc_data.approved.txt

@@ -6,8 +6,9 @@ File type	Conventional base calls
 Encoding	Illumina 1.5
 Total Sequences	5
 Total Bases	80 bp
-Sequences flagged as poor quality	0
 Sequence length	16
+Mean Length	16.0
+Median Length	16
 %GC	50
 >>END_MODULE
 >>Per base sequence quality	pass

test/integration/FileContentsTest_complex_fastqc_report.approved.html

@@ -248,11 +248,15 @@ <h2 id="M0"><img src="data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAACAAAAAgCAYA
        <td>80 bp</td>
       </tr>
       <tr>
-       <td>Sequences flagged as poor quality</td>
-       <td>0</td>
+       <td>Sequence length</td>
+       <td>16</td>
       </tr>
       <tr>
-       <td>Sequence length</td>
+       <td>Mean Length</td>
+       <td>16.0</td>
+      </tr>
+      <tr>
+       <td>Median Length</td>
        <td>16</td>
       </tr>
       <tr>

test/integration/FileContentsTest_minimal_fastqc_data.approved.txt

@@ -6,8 +6,9 @@ File type	Conventional base calls
 Encoding	Illumina 1.5
 Total Sequences	1
 Total Bases	16 bp
-Sequences flagged as poor quality	0
 Sequence length	16
+Mean Length	16.0
+Median Length	16
 %GC	0
 >>END_MODULE
 >>Per base sequence quality	pass

test/integration/FileContentsTest_minimal_fastqc_report.approved.html

@@ -248,11 +248,15 @@ <h2 id="M0"><img src="data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAACAAAAAgCAYA
        <td>16 bp</td>
       </tr>
       <tr>
-       <td>Sequences flagged as poor quality</td>
-       <td>0</td>
+       <td>Sequence length</td>
+       <td>16</td>
       </tr>
       <tr>
-       <td>Sequence length</td>
+       <td>Mean Length</td>
+       <td>16.0</td>
+      </tr>
+      <tr>
+       <td>Median Length</td>
        <td>16</td>
       </tr>
       <tr>

uk/ac/babraham/FastQC/Modules/BasicStats.java

@@ -21,6 +21,7 @@
 
 import java.awt.BorderLayout;
 import java.io.IOException;
+import java.util.Locale;
 
 import javax.swing.JLabel;
 import javax.swing.JPanel;
@@ -249,7 +250,8 @@ public Object getValueAt(int rowIndex, int columnIndex) {
 						}
 
 					case 6 :
-						return ""+(totalBases/actualCount);
+						if (actualCount == 0) return "0.0";
+						return String.format(Locale.ROOT, "%.1f", (double)totalBases/actualCount);
 					case 7 :
 						if (lengthDist == null) {
 							return "NA";

uk/ac/babraham/FastQC/Modules/SequenceLengthDistribution.java

@@ -56,29 +56,37 @@ public boolean ignoreInReport () {
 	}
 
 	public int medianLength() {
-		// We need to go through the data twice - once to get the 
-		// total number of reads and the second time to find the 
-		// length at the 50th percent rank
-
 		long total = 0;
 		for (int i=0;i<lengthCounts.length;i++) {
 			total += lengthCounts[i];
 		}
-
-		long rank50 = total/2;
 
+		if (total == 0) return 0;
+
+		// For an odd number of reads the median is the single central value.
+		// For an even number it's the mean of the two central values, rounded up.
+		if (total % 2 == 1) {
+			return lengthAtRank(total/2);
+		}
+		else {
+			long lo = lengthAtRank((total/2) - 1);
+			long hi = lengthAtRank(total/2);
+			return (int)((lo + hi + 1) / 2);
+		}
+	}
+
+	// Returns the length at the given zero-based rank in ascending order of length.
+	private int lengthAtRank(long rank) {
 		long runningCount = 0;
 		for (int i=0;i<lengthCounts.length;i++) {
-			if (runningCount + lengthCounts[i] > rank50) {
-				return(i);
-			}
-
 			runningCount += lengthCounts[i];
+			if (runningCount > rank) {
+				return i;
+			}
 		}
-
-		throw new RuntimeException("Should never fail to find a value above the 50th percentile");
-	}	
-
+		throw new RuntimeException("Should never fail to find a value above the rank");
+	}
+
 	private synchronized void calculateDistribution () {
 		int maxLen = 0;
 		int minLen = -1;