Changeset 3298

Timestamp:

Aug 8, 2014, 9:15:28 AM (9 years ago)

Author:

Peter

Message:

Yates' correction in Fisher's exact test. fixes #807.

Location:

trunk

Files:

• test/fisher.cc (modified) (4 diffs)
• yat/statistics/Fisher.cc (modified) (4 diffs)
• yat/statistics/Fisher.h (modified) (5 diffs)

trunk/test/fisher.cc

@@ -2 +2 @@
 
 /*
-  Copyright (C) 2008, 2009, 2010, 2012, 2013 Peter Johansson
+  Copyright (C) 2008, 2009, 2010, 2012, 2013, 2014 Peter Johansson
 
   This file is part of the yat library, http://dev.thep.lu.se/yat
@@ -33 +33 @@
 void test_p_value_exact(test::Suite&);
 void test_large_numbers(test::Suite&);
+void test_yates(test::Suite&);
 
 int main(int argc, char* argv[])
@@ -71 +72 @@
   test_p_value(suite);
   test_large_numbers(suite);
+  test_yates(suite);
   return suite.return_value();
 }
@@ -147 +149 @@
   suite.add(suite.equal(f.p_left(), 0.7));
 }
+
+
+void test_yates(test::Suite& suite)
+{
+  statistics::Fisher f;
+  statistics::Fisher f2(true);
+
+  f.oddsratio(5,10,10,10);
+  f2.oddsratio(5,10,10,10);
+  f.minimum_size() = 0;
+  f2.minimum_size() = 0;
+  double p = f.p_value();
+  double p_yates = f2.p_value();
+  if (p==p_yates) {
+    suite.add(false);
+    suite.err() << "p-value: " << p << "\n";
+    suite.err() << "p-value yates's corrected: " << p_yates << "\n";
+  }
+}

trunk/yat/statistics/Fisher.cc

@@ -5 +5 @@
   Copyright (C) 2005 Peter Johansson
   Copyright (C) 2006, 2007, 2008 Jari Häkkinen, Peter Johansson
-  Copyright (C) 2009, 2010, 2011, 2012, 2013 Peter Johansson
+  Copyright (C) 2009, 2010, 2011, 2012, 2013, 2014 Peter Johansson
 
   This file is part of the yat library, http://dev.thep.lu.se/yat
@@ -41 +41 @@
 
 
-  Fisher::Fisher()
-    : a_(0), b_(0), c_(0), d_(0), minimum_size_(10), oddsratio_(1.0)
+  Fisher::Fisher(bool yates)
+    : a_(0), b_(0), c_(0), d_(0), minimum_size_(10), oddsratio_(1.0),
+      yates_(yates)
   {
   }
@@ -62 +63 @@
     double a,b,c,d;
     expected(a,b,c,d);
+    if (yates_)
+      return yates(a_, a) + yates(b_, b) + yates(c_, c) + yates(d_, d);
+
     return (a-a_)*(a-a_)/a + (b-b_)*(b-b_)/b +
       (c-c_)*(c-c_)/c + (d-d_)*(d-d_)/d;
@@ -204 +208 @@
   }
 
+
+  double Fisher::yates(unsigned int o, double e) const
+  {
+    double x = std::abs(o - e) - 0.5;
+    return x*x/e;
+  }
+
 }}} // of namespace statistics, yat, and theplu

trunk/yat/statistics/Fisher.h

@@ -7 +7 @@
   Copyright (C) 2004, 2005 Peter Johansson
   Copyright (C) 2006, 2007, 2008 Jari Häkkinen, Peter Johansson
-  Copyright (C) 2009, 2011, 2013 Peter Johansson
+  Copyright (C) 2009, 2011, 2013, 2014 Peter Johansson
 
   This file is part of the yat library, http://dev.thep.lu.se/yat
@@ -69 +69 @@
     ///
     /// Default Constructor.
-    ///
-    Fisher(void);
+    ///
+    /// \param yates if true Yates's correction is used for
+    /// chi-squared calculation
+    ///
+    Fisher(bool yates=false);
 
     ///
@@ -82 +85 @@
        \frac{(O_i-E_i)^2}{E_i}\f$ where \a E is expected value and \a
        O is observed value.
+
+       If indicated in constructor, Yates's correction is used, i.e.,
+       Chi2 is calculated as \f$ \frac{(|O_i-E_i|-0.5)^2}{E_i} \f$
+
 
        \see expected(double&, double&, double&, double&)
@@ -193 +200 @@
     double p_value_exact(void) const;
 
+    double yates(unsigned int o, double e) const;
+
     unsigned int a_;
     unsigned int b_;
@@ -199 +208 @@
     unsigned int minimum_size_;
     double oddsratio_;
+    bool yates_;
   };