Changeset 372

Timestamp:

Aug 5, 2005, 4:14:02 PM (18 years ago)

Author:

Peter

Message:

modified function normalize in histogram

Location:

trunk/lib/statistics

Files:

• Histogram.cc (modified) (1 diff)
• Histogram.h (modified) (4 diffs)

trunk/lib/statistics/Histogram.cc

@@ -55 +55 @@
 
 
-void Histogram::normalize(double scale_factor)
+void Histogram::normalize(bool choice)
 {
+  double scale_factor;
+  if (choice) 
+    scale_factor = sum_all_.sum_w();
+  else
+    scale_factor = sum_all_.sum_w()*spacing();
   for (size_t i=0; i<histogram_.size(); i++)
     histogram_[i]/=scale_factor;
+
 }
 

trunk/lib/statistics/Histogram.h

@@ -86 +86 @@
 
     ///
-    /// Rescale the histogram using the scale factor \a
-    /// scale_factor. You may need to averager_all or
-    /// averager_histogram to set the proper scale factor.
+    ///  There are two ways to normalize the counts.
     ///
-    /// @todo Redefine the behaviour of this function. The parameter
-    /// should define the integral of the function.
+    /// If choice is true: The normalized count is the count in a
+    /// bin divided by the total number of observations. In this
+    /// case the relative counts are normalized to sum to unity (
+    /// minus values outside histogram).  This is the intuitive case
+    /// where the height of the histogram bar represents the
+    /// proportion of the data in each class.
     ///
-    void normalize(double scale_factor=1);
+    /// If choice is false: The normalized count is the count in the
+    /// class divided by the number of observations times the bin
+    /// width. For this normalization, the area (or integral) under
+    /// the histogram is equal to unity (minus the missing area
+    /// corresponding to counts outside histogram). From a
+    /// probabilistic point of view, this normalization results in a
+    /// relative histogram that is most akin to the probability
+    /// density function If you want to overlay a probability density
+    /// on top of the histogram, use this normalization. Although this
+    /// normalization is less intuitive (relative frequencies greater
+    /// than 1 are quite permissible), it is the appropriate
+    /// normalization if you are using the histogram to model a
+    /// probability density function.
+    ///
+    /// @short Normalizing the histogram
+    /// 
+    void normalize(bool choice = true);
 
     ///
@@ -111 +129 @@
 
     ///
-    /// @return The size of the bins in the histogram.
+    /// @return The width of the bins in the histogram.
     ///
     inline double spacing(void) const { return (xmax_-xmin_)/nof_bins(); }
@@ -118 +136 @@
     /// @return The histogram upper boundary.
     ///
-    /// @note The upper boundary value is inside the histogram.
+    /// @note The upper boundary value is outside the histogram.
     ///
     inline double xmax(void) const { return xmax_; }
@@ -125 +143 @@
     /// @return The histogram lower boundary.
     ///
-    /// @note The lower boundary value is outside the histogram.
+    /// @note The lower boundary value is inside the histogram.
     ///
     inline double xmin(void) const { return xmin_; }
 
     ///
-    /// @return The count of bin \a k.in the histogram.
+    /// @return The count of bin \a k in the histogram.
+    ///
     inline double operator[](size_t k) const { return histogram_[k]; } 
 
     ///
     /// The assignment operator
+    ///
     const Histogram& operator=(const Histogram&);