sortedClassCount = sorted(classCount.iteritems(), key=operator.itemgetter(1), reverse=True) # Sort classCount (highest to lowest) by voteIlabel count
AttributeError: 'dict' object has no attribute 'iteritems'
v2.0 grammar:
sortedClassCount = sorted(classCount.iteritems(), key = operator.itemgetter(1), reverse = True) #2.0
v3.0 grammar:
sortedClassCount = sorted(classCount.items(), key = operator.itemgetter(1), reverse = True) #3.0
code:
# Use k Nearest Neighbors to classify inX accortding to existing dataSet with known ratings in labels
def classify0(inX, dataSet, labels, k):
dataSetSize = dataSet.shape[0] # Number of entries in dataSet
diffMat = tile(inX, (dataSetSize,1)) - dataSet # Array of same shape as dataSet holding inX-dataSet entry in each position
sqDiffMat = diffMat**2 # Square entries in diffMat
sqDistances = sqDiffMat.sum(axis=1) # Sum over vector components
distances = sqDistances**0.5 # Traditional Euclidean distance foreach dataSet entry
sortedDistIndicies = distances.argsort() # argsort returns indices that sort distances
classCount={} # An empty dictionary key:value pairs key = labels value = number times this label in set of k
for i in range(k): # Run over k nearest neighbors
voteIlabel = labels[sortedDistIndicies[i]] # Label of this neighbor
classCount[voteIlabel] = classCount.get(voteIlabel,0) + 1 # .get gets current count for voteIlabel and returns zero if first time voteIlabel seen (default = 0)
sortedClassCount = sorted(classCount.items(), key=operator.itemgetter(1), reverse=True) # Sort classCount (highest to lowest) by voteIlabel count
return sortedClassCount[0][0] # The label that occurs most often in k nearest neighbors