StatRed ass3: Changed file structures.

statred/ass3/Makefile

+SEED=100
+
+knnb:
+	echo 'nnb:'; \
+	python classify.py nnb ${SEED}; \
+	for i in 1 3 5 7 9; do \
+		echo 'knnb '$$i':'; \
+		python classify.py knnb $$i ${SEED}; \
+	done
+
+clean:
+	rm knnb-*.pdf

statred/ass3/classifiers.py

+from pylab import argmin, tile
+
+class NNb:
+    def __init__(self, X, c):
+        self.n, self.N = X.shape
+        self.X, self.c = X, c
+
+    def classify(self, x):
+        d = self.X - tile(x.reshape(self.n, 1), self.N);
+        dsq = sum(d*d, 0)
+        return self.c[argmin(dsq)]
+
+class kNNb:
+    def __init__(self, X, c, k):
+        self.n, self.N = X.shape
+        self.X, self.c, self.k = X, c, k
+
+    def classify(self, x):
+        d = self.X - tile(x.reshape(self.n, 1), self.N);
+        dsq = sum(d*d, 0)
+        minindices = dsq.argsort()
+        # Count class occurrences in k nearest neighbours
+        hist = {}
+        for c in self.c[minindices[:self.k]]:
+            try:
+                hist[c] += 1
+            except KeyError:
+                hist[c] = 1
+        # Return the majority class
+        max_nbb = (0, None)
+        for c, count in hist.iteritems():
+            if count > max_nbb[0]:
+                max_nnb = (count, c)
+        return max_nnb[1]
+
+class MEC:
+    def __init__(self, X, c):
+        self.n, self.N = X.shape
+        self.X, self.c = X, c
+
+    def classify(self, x):
+        # TODO: Implement MEC classifier
+        pass

statred/ass3/1-knnb.py → statred/ass3/classify.py

 from pylab import loadtxt, arange, loadtxt, permutation, transpose,\
     zeros, sum, plot, subplot, array, scatter, logical_and, figure,\
-    show, savefig, tile, argmin, seed
+    savefig, seed
 from sys import argv, exit
+import classifiers
 
-try:
-    k = int(argv[1])
-except IndexError:
-    print 'Usage: python %s K [ SEED ] (use K = -1 for the regular NNb method)' % argv[0]
+# Parse input
+argc = len(argv)
+if argc < 2:
+    print 'Usage: python %s METHOD [ K ] [ SEED ]' % argv[0]
     exit()
-
-try:
+method = argv[1]
+if method == 'knnb':
+    if argc < 3:
+        print 'Usage: python %s %s K [ SEED ]' % tuple(argv[:2])
+        exit()
+    k = int(argv[2])
+    if argc == 4:
+        seed(int(argv[3]))
+elif method not in ['nnb', 'mec']:
+    print 'Unknown classification method "%s"' % argv[1]
+    exit()
+elif argc == 3:
     seed(int(argv[2]))
-except IndexError:
-    pass
-
-class NNb:
-    def __init__(self, X, c, k):
-        self.n, self.N = X.shape
-        self.X, self.c = X, c
-
-    def classify(self, x):
-        d = self.X - tile(x.reshape(self.n, 1), self.N);
-        dsq = sum(d*d, 0)
-        return self.c[argmin(dsq)]
-
-class kNNb:
-    def __init__(self, X, c, k):
-        self.n, self.N = X.shape
-        self.X, self.c, self.k = X, c, k
-
-    def classify(self, x):
-        d = self.X - tile(x.reshape(self.n, 1), self.N);
-        dsq = sum(d*d, 0)
-        minindices = dsq.argsort()
-        # Count class occurrences in k nearest neighbours
-        hist = {}
-        for c in self.c[minindices[:self.k]]:
-            try:
-                hist[c] += 1
-            except KeyError:
-                hist[c] = 1
-        # Return the majority class
-        max_nbb = (0, None)
-        for c, count in hist.iteritems():
-            if count > max_nbb[0]:
-                max_nnb = (count, c)
-        return max_nnb[1]
 
+# Load dataset
 def cnvt(s):
     tab = {'Iris-setosa': 1.0, 'Iris-versicolor': 2.0, 'Iris-virginica': 3.0}
     if tab.has_key(s):
@@ -60,26 +37,31 @@ ind = permutation(ind) # random permutation
 L = ind[0:90] # learning set indices
 T = ind[90:] # test set indices
 
-# Learning Set
+# Learning set
 X = transpose(XC[L,0:4])
-if k == -1:
-    knnc = NNb(X, XC[L,-1], k)
-else:
-    knnc = kNNb(X, XC[L,-1], k)
+args = [X, XC[L,-1]]
+if method == 'nnb':
+    method_class = classifiers.NNb
+elif method == 'knnb':
+    method_class = classifiers.kNNb
+    args.append(k)
+elif method == 'mec':
+    method_class = classifiers.MEC
+classifier = method_class(*args)
 
-# Classification of Test Set
+# Classification of test set
 c = zeros(len(T))
 for i in arange(len(T)):
-    c[i] = knnc.classify(XC[T[i],0:4])
+    c[i] = classifier.classify(XC[T[i],0:4])
 
-# Confusion Matrix
+# Confusion matrix
 CM = zeros((3, 3))
 for i in range(3):
     for j in range(3):
         CM[i,j] = sum(logical_and(XC[T,4] == i + 1, c == j + 1))
 print CM
 
-# Plot Test Set
+# Plot test set
 figure(1)
 color = array([[1,0,0], [0,1,0], [0,0,1]])
 for i in range(4):
@@ -92,5 +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()
-savefig('knnb-%d.pdf' % k)
+if method == 'knnb':
+    filename = 'knnb-%d.pdf' % k
+else:
+    filename = '%s.pdf' % method
+savefig(filename)