From 92280b4068e80bf17b6d4a615e6604a6ad6e8be1 Mon Sep 17 00:00:00 2001
From: Eric Banks <ebanks@broadinstitute.org>
Date: Fri, 8 Jun 2012 13:54:37 -0400
Subject: [PATCH] BQSR optimization: cache the BitSetUtils.bitSetFrom() calls
 since they are called over and over again with the same values.  Another 10%
 reduction in runtime.

---
 .../gatk/walkers/bqsr/BQSRKeyManager.java     | 21 ++++++++++++++-----
 .../sting/utils/BitSetUtils.java              | 11 +++++++++-
 2 files changed, 26 insertions(+), 6 deletions(-)

diff --git a/public/java/src/org/broadinstitute/sting/gatk/walkers/bqsr/BQSRKeyManager.java b/public/java/src/org/broadinstitute/sting/gatk/walkers/bqsr/BQSRKeyManager.java
index c1449c5b3..728d1a8d4 100644
--- a/public/java/src/org/broadinstitute/sting/gatk/walkers/bqsr/BQSRKeyManager.java
+++ b/public/java/src/org/broadinstitute/sting/gatk/walkers/bqsr/BQSRKeyManager.java
@@ -66,7 +66,7 @@ public class BQSRKeyManager {
         for (Covariate optional : optionalCovariates) {
             int nBits = optional.numberOfBits();                                                                        // number of bits used by this covariate
             nOptionalBits = Math.max(nOptionalBits, nBits);                                                             // optional covariates are represented by the number of bits needed by biggest covariate
-            BitSet optionalID = BitSetUtils.bitSetFrom(id);                                                             // calculate the optional covariate ID for this covariate
+            BitSet optionalID = bitSetFromId(id);                                                                       // calculate the optional covariate ID for this covariate
             optionalCovariatesInfo.add(new OptionalCovariateInfo(optionalID, optional));                                // optional covariates have standardized mask and number of bits, so no need to store in the RequiredCovariateInfo object
             String covariateName = optional.getClass().getSimpleName().split("Covariate")[0];                           // get the name of the covariate (without the "covariate" part of it) so we can match with the GATKReport
             covariateNameToIDMap.put(covariateName, id);
@@ -100,7 +100,7 @@ public class BQSRKeyManager {
     public List<BitSet> bitSetsFromAllKeys(BitSet[] allKeys, EventType eventType) {
         List<BitSet> allBitSets = new ArrayList<BitSet>();                                                              // Generate one key per optional covariate
 
-        BitSet eventBitSet = BitSetUtils.bitSetFrom(eventType.index);                                                   // create a bitset with the event type
+        BitSet eventBitSet = bitSetFromEvent(eventType);                                                                // create a bitset with the event type
         int eventTypeBitIndex = nRequiredBits + nOptionalBits + nOptionalIDBits;                                        // Location in the bit set to add the event type bits
 
         int covariateIndex = 0;
@@ -257,9 +257,20 @@ public class BQSRKeyManager {
             eventKey.set(i - firstBitIndex);
         return EventType.eventFrom(BitSetUtils.shortFrom(eventKey));
     }
-    
-    private BitSet bitSetFromEvent(EventType eventType) {
-        return BitSetUtils.bitSetFrom(eventType.index);
+
+    // cache the BitSet representing an event since it's otherwise created a massive amount of times
+    private static final Map<EventType, BitSet> eventTypeCache = new HashMap<EventType, BitSet>(EventType.values().length);
+    static {
+        for (final EventType eventType : EventType.values())
+            eventTypeCache.put(eventType, BitSetUtils.bitSetFrom(eventType.index));
+    }
+
+    private BitSet bitSetFromEvent(final EventType eventType) {
+        return eventTypeCache.get(eventType);
+    }
+
+    private BitSet bitSetFromId(final short id) {
+        return BitSetUtils.bitSetFrom(id);
     }
 
     private int bitsInEventType() {
diff --git a/public/java/src/org/broadinstitute/sting/utils/BitSetUtils.java b/public/java/src/org/broadinstitute/sting/utils/BitSetUtils.java
index 41889e2f5..aa7cd4b37 100644
--- a/public/java/src/org/broadinstitute/sting/utils/BitSetUtils.java
+++ b/public/java/src/org/broadinstitute/sting/utils/BitSetUtils.java
@@ -5,6 +5,8 @@ import org.broadinstitute.sting.utils.exceptions.ReviewedStingException;
 import java.io.ByteArrayOutputStream;
 import java.io.ObjectOutputStream;
 import java.util.BitSet;
+import java.util.HashMap;
+import java.util.Map;
 
 /**
  * Utilities for bitset conversion
@@ -71,8 +73,15 @@ public class BitSetUtils {
      * @return a bitset representation of the short
      */
     public static BitSet bitSetFrom(short number) {
-        return bitSetFrom(number, NBITS_SHORT_REPRESENTATION);
+        BitSet result = shortCache.get(number);
+        if (result == null) {
+            result = bitSetFrom(number, NBITS_SHORT_REPRESENTATION);
+            shortCache.put(number, result);
+        }
+        return result;
     }
+    // use a static cache for shorts (but not for longs, because there could be a lot of entries)
+    private static final Map<Short, BitSet> shortCache = new HashMap<Short, BitSet>(2 * Short.MAX_VALUE);
 
     /**
      * Creates a BitSet representation of an arbitrary integer (number of bits capped at 64 -- long precision)