StatRed ass3: Implemented part 2.

statred/ass3/classifiers.py

-from pylab import argmin, tile
+from pylab import argmin, argmax, tile, unique, argwhere, array, mean, \
+        newaxis, dot, pi, e, matrix
 
 class NNb:
     def __init__(self, X, c):
@@ -37,7 +38,24 @@ class MEC:
     def __init__(self, X, c):
         self.n, self.N = X.shape
         self.X, self.c = X, c
+        self.classes = unique(c)
+        self.estimate()
+
+    def estimate(self):
+        """Estimate the mean and covariance matrix each class in the learning
+        set."""
+        self.class_data = []
+        for c in self.classes:
+            indices = argwhere(array(map(lambda x: x if x == c else 0,
+                self.c))).T[0]
+            X = self.X[:,indices]
+            mu = mean(X, 1)
+            Yzm = X - tile(mu[:,newaxis], X.shape[1])
+            S = matrix(dot(Yzm, Yzm.T) / (self.n - 1))
+            coeff = 1 / (S.A**-.5 * (2 * pi)**(self.n / 2))
+            self.class_data.append((mu, S, coeff))
 
     def classify(self, x):
-        # TODO: Implement MEC classifier
-        pass
+        p = [coeff * e**(-.5 * dot(x - mu, dot(S.I, array([x -
+            mu]).T)).tolist()[0][0]) for mu, S, coeff in self.class_data]
+        return self.classes[argmax([i.sum() for i in p])]

statred/ass3/classify.py

@@ -38,7 +38,7 @@ L = ind[0:90] # learning set indices
 T = ind[90:] # test set indices
 
 # Learning set
-X = transpose(XC[L,0:4])
+X = XC[L,0:4]
 args = [X, XC[L,-1]]
 if method == 'nnb':
     method_class = classifiers.NNb
@@ -74,10 +74,10 @@ for i in range(4):
             facecolor=[1,1,1]*len(T))
         scatter(XC[T, i], XC[T, j], s=30, marker='+',
             edgecolor=color[c.astype(int) - 1])
-#from pylab import show
-#show()
-if method == 'knnb':
-    filename = 'knnb-%d.pdf' % k
-else:
-    filename = '%s.pdf' % method
-savefig(filename)
+from pylab import show
+show()
+#if method == 'knnb':
+#    filename = 'knnb-%d.pdf' % k
+#else:
+#    filename = '%s.pdf' % method
+#savefig(filename)