Return-Path: 
 <commits-return-2243-apmail-ctakes-commits-archive=ctakes.apache.org@ctakes.apache.org>
X-Original-To: apmail-ctakes-commits-archive@www.apache.org
Delivered-To: apmail-ctakes-commits-archive@www.apache.org
Received: from mail.apache.org (hermes.apache.org [140.211.11.3])
	by minotaur.apache.org (Postfix) with SMTP id 91BECC7C6
	for <apmail-ctakes-commits-archive@www.apache.org>;
 Fri, 19 Jul 2013 16:35:38 +0000 (UTC)
Received: (qmail 45510 invoked by uid 500); 19 Jul 2013 16:35:38 -0000
Delivered-To: apmail-ctakes-commits-archive@ctakes.apache.org
Received: (qmail 45449 invoked by uid 500); 19 Jul 2013 16:35:35 -0000
Mailing-List: contact commits-help@ctakes.apache.org; run by ezmlm
Precedence: bulk
List-Help: <mailto:commits-help@ctakes.apache.org>
List-Unsubscribe: <mailto:commits-unsubscribe@ctakes.apache.org>
List-Post: <mailto:commits@ctakes.apache.org>
List-Id: <commits.ctakes.apache.org>
Reply-To: dev@ctakes.apache.org
Delivered-To: mailing list commits@ctakes.apache.org
Received: (qmail 45427 invoked by uid 99); 19 Jul 2013 16:35:33 -0000
Received: from athena.apache.org (HELO athena.apache.org) (140.211.11.136)
    by apache.org (qpsmtpd/0.29) with ESMTP; Fri, 19 Jul 2013 16:35:33 +0000
X-ASF-Spam-Status: No, hits=-2000.0 required=5.0
	tests=ALL_TRUSTED
X-Spam-Check-By: apache.org
Received: from [140.211.11.4] (HELO eris.apache.org) (140.211.11.4)
    by apache.org (qpsmtpd/0.29) with ESMTP; Fri, 19 Jul 2013 16:35:30 +0000
Received: from eris.apache.org (localhost [127.0.0.1])
	by eris.apache.org (Postfix) with ESMTP id BAFA3238889B;
	Fri, 19 Jul 2013 16:35:10 +0000 (UTC)
Content-Type: text/plain; charset="utf-8"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Subject: svn commit: r1504935 -
 /ctakes/trunk/ctakes-temporal/src/main/java/org/apache/ctakes/temporal/ae/feature/selection/Chi2FeatureSelection.java
Date: Fri, 19 Jul 2013 16:35:10 -0000
To: commits@ctakes.apache.org
From: clin@apache.org
X-Mailer: svnmailer-1.0.9
Message-Id: <20130719163510.BAFA3238889B@eris.apache.org>
X-Virus-Checked: Checked by ClamAV on apache.org

Author: clin
Date: Fri Jul 19 16:35:10 2013
New Revision: 1504935

URL: http://svn.apache.org/r1504935
Log:
Make a Yate's parameter for Chi2 feature selection. So people may turn on or off the Yate's correction for Chi2 value calculation.
If Yate's correction is on (boolean Yates = true), small difference between observed value and expected value (<0.5) will be ignored. More features will be trimmed.
Else, small difference will be kept. Users have the freedom to keep all features. 

Modified:
    ctakes/trunk/ctakes-temporal/src/main/java/org/apache/ctakes/temporal/ae/feature/selection/Chi2FeatureSelection.java

Modified: ctakes/trunk/ctakes-temporal/src/main/java/org/apache/ctakes/temporal/ae/feature/selection/Chi2FeatureSelection.java
URL: http://svn.apache.org/viewvc/ctakes/trunk/ctakes-temporal/src/main/java/org/apache/ctakes/temporal/ae/feature/selection/Chi2FeatureSelection.java?rev=1504935&r1=1504934&r2=1504935&view=diff
==============================================================================
--- ctakes/trunk/ctakes-temporal/src/main/java/org/apache/ctakes/temporal/ae/feature/selection/Chi2FeatureSelection.java (original)
+++ ctakes/trunk/ctakes-temporal/src/main/java/org/apache/ctakes/temporal/ae/feature/selection/Chi2FeatureSelection.java Fri Jul 19 16:35:10 2013
@@ -38,9 +38,12 @@ public class Chi2FeatureSelection<OUTCOM
 
     protected Table<String, OUTCOME_T, Integer> featValueClassCount;
 
-    public Chi2Scorer() {
+	private boolean yates = false;
+
+    public Chi2Scorer(boolean yate) {
       this.classCounts = HashMultiset.<OUTCOME_T> create();
       this.featValueClassCount = HashBasedTable.<String, OUTCOME_T, Integer> create();
+      this.yates = yate;
     }
 
     public void update(String featureName, OUTCOME_T outcome, int occurrences) {
@@ -88,13 +91,12 @@ public class Chi2FeatureSelection<OUTCOM
         return chi2val;
       }
 
-      boolean yates = true;
       for (int lbl = 0; lbl < numOfClass; lbl++) {
         // for positive part of feature:
         double expected = (outcomeCounts[lbl] / (double) n) * (posiFeatCount);
         if (expected > 0) {
           double diff = Math.abs(posiOutcomeCounts[lbl] - expected);
-          if (yates) { // apply Yate's correction
+          if (this.yates ) { // apply Yate's correction
             diff -= 0.5;
           }
           if (diff > 0)
@@ -106,7 +108,7 @@ public class Chi2FeatureSelection<OUTCOM
         double observ = outcomeCounts[lbl] - posiOutcomeCounts[lbl];
         if (expected > 0) {
           double diff = Math.abs(observ - expected);
-          if (yates) { // apply Yate's correction
+          if (this.yates) { // apply Yate's correction
             diff -= 0.5;
           }
           if (diff > 0)
@@ -121,6 +123,8 @@ public class Chi2FeatureSelection<OUTCOM
   private double chi2Threshold;
 
   private Chi2Scorer<OUTCOME_T> chi2Function;
+  
+  private boolean yates = false;
 
   public Chi2FeatureSelection(String name) {
     this(name, 0.0);
@@ -131,6 +135,17 @@ public class Chi2FeatureSelection<OUTCOM
     this.chi2Threshold = threshold;
   }
 
+  /**
+   * Constructor that can let use control the yate's correction
+   * @param name
+   * @param threshold
+   * @param yates : true for using yate's correction, false for turn off yate's correction
+   */
+  public Chi2FeatureSelection(String name, double threshold, boolean yates) {
+	    super(name);
+	    this.chi2Threshold = threshold;
+	    this.yates = yates;
+	  }
   @Override
   public boolean apply(Feature feature) {
     return this.selectedFeatureNames.contains(this.getFeatureName(feature));
@@ -139,7 +154,7 @@ public class Chi2FeatureSelection<OUTCOM
   @Override
   public void train(Iterable<Instance<OUTCOME_T>> instances) {
     // aggregate statistics for all features
-    this.chi2Function = new Chi2Scorer<OUTCOME_T>();
+    this.chi2Function = new Chi2Scorer<OUTCOME_T>(this.yates);
     for (Instance<OUTCOME_T> instance : instances) {
       OUTCOME_T outcome = instance.getOutcome();
       for (Feature feature : instance.getFeatures()) {