def classification(self, k_value, length, actual_z1, actual_z2):
predicitions = []
for x in range(length):
neighbors = knn_func.getNeighbors(actual_z1, actual_z2[x], k_value)
result = knn_func.getResponse(neighbors)
predicitions.append(result)
accuracy = knn_func.getAccuracy(self.dataset_z2, predicitions)
knn_func.createErrorFile(self.dataset_z2, predicitions)
return accuracy
def bestZ1(k_value):
i = 1
predictions = []
accuracy_list = []
while (i <= 10):
instance = "z1n{0}.txt".format(i)
print("Testing instance: {0}".format(instance))
with open(instance, 'r') as content:
lines = csv.reader(content)
dataset_z1 = list(lines)
length_z1 = len(dataset_z1)
for x in range(len(dataset_z1)):
for y in range(4):
dataset_z1[x][y] = float(dataset_z1[x][y])
with open('z2.txt', 'r') as content:
lines = csv.reader(content)
dataset_z2 = list(lines)
length_z2 = len(dataset_z2)
for x in range(length_z2):
for y in range(4):
dataset_z2[x][y] = float(dataset_z2[x][y])
for x in range(length_z1):
neighbors = knn_func.getNeighbors(dataset_z1, dataset_z2[x],
k_value)
result = knn_func.getResponse(neighbors)
predictions.append(result)
accuracy = knn_func.getAccuracy(dataset_z2, predictions)
print("accuracy: ", accuracy)
accuracy_list.append(accuracy)
predictions = []
i = i + 1
maximum = 0
for x in range(len(accuracy_list)):
if accuracy_list[x] > maximum:
maximum = accuracy_list[x]
best_i = x + 1
print("Best instance of Z1 is: z1n{0}, with {1}%% of approval".format(
best_i, maximum))
return best_i
def bestK():
with open('z1.txt', 'r') as content:
lines = csv.reader(content)
dataset_z1 = list(lines)
length_z1 = len(dataset_z1)
for x in range(len(dataset_z1)):
for y in range(4):
dataset_z1[x][y] = float(dataset_z1[x][y])
with open('z2.txt', 'r') as content:
lines = csv.reader(content)
dataset_z2 = list(lines)
length_z2 = len(dataset_z2)
for x in range(length_z2):
for y in range(4):
dataset_z2[x][y] = float(dataset_z2[x][y])
k = 1
limit = math.trunc(length_z1 / 2)
predictions = []
accuracy_list = []
k_list = []
while (k < limit):
print("Testing K = ", k)
for x in range(0, length_z1):
neighbors = knn_func.getNeighbors(dataset_z1, dataset_z2[x], k)
result = knn_func.getResponse(neighbors)
predictions.append(result)
accuracy = knn_func.getAccuracy(dataset_z2, predictions)
knn_func.createErrorFile(dataset_z2, predictions)
accuracy_list.append(accuracy)
predictions = []
print('Accuracy: ' + str(accuracy) + '%')
k_list.append(k)
k = k + 2
maximum = 0
for x in range(0, len(accuracy_list)):
if accuracy_list[x] > maximum:
maximum = accuracy_list[x]
best_k = k_list[x]
print("Best K is: {0} with {1}%%".format(best_k, maximum))
return (best_k, maximum)
def classifyZ3(k_value, instance_z1):
k_value = int(k_value)
with open(instance_z1, 'r') as content:
lines = csv.reader(content)
dataset_z1 = list(lines)
length_z1 = len(dataset_z1)
for x in range(length_z1):
for y in range(4):
dataset_z1[x][y] = float(dataset_z1[x][y])
with open('z3.txt', 'r') as content:
lines = csv.reader(content)
dataset_z3 = list(lines)
length_z3 = len(dataset_z3)
for x in range(length_z3):
for y in range(4):
dataset_z3[x][y] = float(dataset_z3[x][y])
predictions = []
accuracy_list = []
for x in range(length_z3):
neighbors = knn_func.getNeighbors(dataset_z1, dataset_z3[x], k_value)
result = knn_func.getResponse(neighbors)
predictions.append(result)
accuracy = knn_func.getAccuracy(dataset_z3, predictions)
print("accuracy: ", accuracy)
accuracy_list.append(accuracy)
predictions = []
maximum = 0
for x in range(len(accuracy_list)):
if accuracy_list[x] > maximum:
maximum = accuracy_list[x]
print("Result Z3 classification with K = {0}, using {1} instance of Z1, presents {2}%% efficiency".format(k_value, instance_z1, maximum))
return maximum
def main():
# READ IRIS && TRAIN/TEST SPLIT
trainingSet=[]
testSet=[]
split = 0.7
loadDataset('iris.data', split, trainingSet, testSet)
print ('Train set: ' + repr(len(trainingSet)))
print ('Test set: ' + repr(len(testSet)))
# NEIGHBOURS CALCULATION && RESPONSE PREDICTION
predictions=[]
k = 3
for x in range(len(testSet)):
neighbors = getNeighbors(trainingSet, testSet[x], k)
predicted = getResponse(neighbors)
predictions.append(predicted)
print('> predicted=' + repr(predicted) + ', actual=' + repr(testSet[x][-1]))
# ACCURACY ESTIMATION
accuracy = getAccuracy(testSet, predictions)
print('Accuracy: ' + repr(accuracy) + '%')