Improved comment explaining numerical stability issue

Author: mnielsen

code/network2.py

@@ -43,9 +43,11 @@ class CrossEntropyCost:
     @staticmethod
     def fn(a, y):
         """Return the cost associated with an output ``a`` and desired output
-        ``y``.  Note that the np.nan_to_num ensures that if the output
-        from the network is exactly right, then 0.0 will be returned,
-        rather than nan.
+        ``y``.  Note that np.nan_to_num is used to ensure numerical
+        stability.  In particular, if both ``a`` and ``y`` have a 1.0
+        in the same slot, then the expression (1-y)*np.log(1-a)
+        returns nan.  The np.nan_to_num ensures that that is converted
+        to the correct value (0.0).
 
         """
         return np.nan_to_num(np.sum(-y*np.log(a)-(1-y)*np.log(1-a)))