Adding a kNN classifier description.

openAccess/gitPLoS/classifiers/kNN.py

@@ -1,11 +1,24 @@
 #!/usr/bin/env python
+# openAccess: getPLoS
 #
+#Copyright (c) 2001-2012 openAccess Project
+# Author: Bill OConnor
+# URL: <https://github.com/openAccess/gitPLoS>
+# For license information, see LICENSE.TXT
 #
 """
+    Description
+    ===========
+    
+    kNN - k Nearest Neighbor classifier. Takes a list of 
+    (category, [values]) tuples as training data. Data points
+    are classified by calculating the distance from each 
+    point in the training data. The distances are sorted giving
+    higher preference to smallest distances. Then k shortest
+    distances are used to calculate the weights for each class.
 """
 from __future__ import division
-import math
-from collections import *
+from math import sqrt
 
 def _vec_minkowski(p, q, e):
     l = [ abs(p[i] - q[i])**e for i in xrange(len(q)) ]
@@ -19,10 +32,10 @@ def _vec_manhattan_dist(p, q):
 
 def _vec_euclidean_dist(p, q):
     """
-    Vector distance function. 
+    Euclidean Vector distance function. 
     """
     l = [ p[i]-q[i] for i in xrange(len(p)) ]
-    return math.sqrt(sum([ d*d for d in l ]))
+    return sqrt(sum([ d*d for d in l ]))
 
 def _euclidean_dist(x, y):
     """
@@ -35,10 +48,11 @@ def _eq_weight(x, y):
 
 class kNN(object):
     """
+    k Nearest Neighbor classifier.
     """
     def __init__(self, data, k):
         """
-	data - a list of (category, value) tuples.
+	data - a list of (category, [values] ) tuples.
         """
         self.data = data    # list of (category, value) tuples used for training
         self.k = k
@@ -67,11 +81,3 @@ def classify(self, x, **kwargs):
 	if len(class_lst) > 0:
             return class_lst[0]
         return None
-
-if __name__ == '__main__':
-    data = [ ('a', [1.0]), ('b', [3.0]), ('b', [4.0]), ('a', [1.5]), ('c', [3.5]) ]
-    knn = kNN(data,3)
-    r = knn.calculate([4.0])
-    print knn.classify([4.0])
-    print r
-