def test_accuracy(rootNode):
print("Testing Accuracy of the Algorithm")
testCptTable = parse_training_data('data/featnames.csv','data/testfeat.csv','data/testlabs.csv')
num_corrects = 0
num_wrongs = 0
for row in testCptTable:
#for each row test the accuracy
if not IsCompliant(row, rootNode):
num_wrongs+=1
else:
num_corrects+=1
total_count = num_corrects+num_wrongs
accuracy = (num_corrects/total_count)*100
print(" Accuracy Percentage ="+str(accuracy))
def test_accuracy(rootNode):
print("Testing Accuracy of the Algorithm")
testCptTable = parse_training_data('data/featnames.csv',
'data/testfeat.csv',
'data/testlabs.csv')
num_corrects = 0
num_wrongs = 0
for row in testCptTable:
#for each row test the accuracy
if not IsCompliant(row, rootNode):
num_wrongs += 1
else:
num_corrects += 1
total_count = num_corrects + num_wrongs
accuracy = (num_corrects / total_count) * 100
print(" Accuracy Percentage =" + str(accuracy))
def main():
print(" Please choose if you want to use chi-square:")
print(" 1. Do not use Chi-Square")
print(" 2. Use Chi-Square")
userInput = input(" Enter your choice:")
global p_value
useChiSq = False
if userInput == 1:
useChiSq = False
elif userInput == 2:
userPvalue = input(" Enter p value, Valid values (0.01|0.05|1):")
if (userPvalue != 0.01) and (userPvalue != 0.05) and (userPvalue != 1):
print(" Please enter a valid P value, exiting")
return
p_value = str(userPvalue)
useChiSq = True
else:
print(" Please make a valid choice, exiting ")
return
global total_size
total_size = 0
print(" Given p value " + str(p_value))
#CPT table is a list of rows with attributes and values, approximately of 40k size
print(" Parsing Training Data")
cptTable = parse_training_data('data/featnames.csv', 'data/trainfeat.csv',
'data/trainlabs.csv')
global total_num_rows
total_num_rows = len(cptTable)
attributes = cptTable[0].keys()
attributes.remove('nextPage')
print(" Building decision tree")
root_node = build_id3(cptTable, attributes, useChiSq)
#bestInfoGainAttr(cptTable)
#print_id3(root_node)
test_accuracy(root_node)
calculatespace(root_node)
print("Total Size (Nodes in decision tree)=" + str(total_size))
def main():
print(" Please choose if you want to use chi-square:")
print(" 1. Do not use Chi-Square")
print(" 2. Use Chi-Square")
userInput = input(" Enter your choice:")
global p_value
useChiSq = False
if userInput == 1:
useChiSq = False
elif userInput == 2:
userPvalue = input(" Enter p value, Valid values (0.01|0.05|1):")
if (userPvalue != 0.01) and (userPvalue != 0.05) and (userPvalue != 1):
print(" Please enter a valid P value, exiting")
return
p_value = str(userPvalue)
useChiSq = True
else:
print(" Please make a valid choice, exiting ")
return
global total_size
total_size =0
print(" Given p value "+str(p_value))
#CPT table is a list of rows with attributes and values, approximately of 40k size
print(" Parsing Training Data")
cptTable = parse_training_data('data/featnames.csv','data/trainfeat.csv','data/trainlabs.csv')
global total_num_rows
total_num_rows = len(cptTable)
attributes = cptTable[0].keys()
attributes.remove('nextPage')
print(" Building decision tree")
root_node = build_id3(cptTable, attributes, useChiSq)
#bestInfoGainAttr(cptTable)
#print_id3(root_node)
test_accuracy(root_node)
calculatespace(root_node)
print("Total Size (Nodes in decision tree)="+str(total_size))
