def DTfunc(arg1, arg2, arg3=None):
#encode tree-depths
if "balance.scale" in arg1:
depth = 3
if "nursery" in arg1:
depth = 7
if "led" in arg1:
depth = 7
if "synthetic.social" in arg1:
depth = 8
#Read training file
Rf = ReadTrFile.ReadTrFile(arg1)
#Read Testing file
Tf = ReadTeFile.ReadTeFile(arg2)
#Initializer the Trainer class
Train = Training.Training(depth, Rf.D, Rf.dictfile, Rf.attributedict,
Rf.labeldict, Rf.avcdict)
#Generate Decision Tree
trlabeldict, dtroot = Train.GenDT()
#Initialize Tester
Test = Testing.Testing(dtroot, Tf.dictfile)
#Perform Decision Tree based Testing
prlabel = Test.parsetestfile()
#Calculate and Print Confusion Matrix
cm = utilities.CalcConMat(Tf.dictfile, prlabel, Rf.labeldict)
#Calculate quality of classifer for report
if arg3 == True:
QualityCalc.QualityCalc(cm)
def RTfunc(arg1, arg2, arg3=None):
#encode num-trees
if "balance.scale" in arg1:
numtrees = 150
numatt = 2
depth = 2
databag = 1
if "nursery" in arg1:
numtrees = 100
numatt = 8
depth = 7
databag = 1
if "led" in arg1:
numtrees = 65
numatt = 7
depth = 7
databag = 1
if "synthetic.social" in arg1:
numtrees = 135
numatt = 12
depth = 8
databag = 1
#Read Training file
Rf = ReadTrFile.ReadTrFile(arg1)
#Read Testing file
Tf = ReadTeFile.ReadTeFile(arg2)
#Build DT's and test them
prlabel = {}
for x in range(0, numtrees):
Dr = []
size = len(Rf.D)
for y in range(0, size):
#Sampling with replacement
if databag == 1:
index = random.randint(0, size - 1)
else:
index = y
Dr.append(Rf.D[index])
Train = Training.Training(depth, Dr, Rf.dictfile, Rf.attributedict)
dtroot = Train.GenDTRF(numatt)
Test = Testing.Testing(dtroot, Tf.dictfile, prlabel)
prlabel = Test.parsetestfile()
#Calculate Confusion Matrix for Random forest results
cm = utilities.CalcConMat(Tf.dictfile, prlabel, Rf.labeldict)
#Calculate quality
if arg3 == True:
QualityCalc.QualityCalc(cm)
def main():
arguments = parse_arguments()
if arguments.test:
Testing.Testing().run_tests()
else:
# process input file and store the events read
events = utility.process_input_data(arguments.input_file)
# create moving average class with event and window size
moving_average = MovingAverage.MovingAverage(events, arguments.window)
# perform moving average time calculation
average_times = moving_average.moving_average()
# for output in average_times:
utility.log_average_time(arguments.input_file, arguments.print,
average_times, arguments.output_file)
import Testing
import Training
# /////////////////////////////////////////////////////////////////
x = float(input("Enter 1 To train Or 2 To Test : "))
if x == 1:
train = Training.Training()
train.training()
else:
c = Testing.Testing()
c.testing()
plt.title("Report")
iteration = []
dt, dl, td = Utils.Utils(dt, dl, td).orderData(3)
seper = 5
testArgs = range(seper)
w_per = [[], [], [], [], []]
w_pa = [[], [], [], [], []]
w_svm = [[], [], [], [], []]
tester = [Training, Training, Training, Training, Training]
for tr in [0]:
for i in testArgs:
i = int(i)
iteration.append(i)
w_per[i], w_pa[i], w_svm[i] = Training.Training(dt, dl,
params).train(i)
tester[i] = Testing.Testing(dt, dl, w_per[i], w_pa[i], w_svm[i])
if len(sys.argv) == 3: # debug mode
t1, t2, t3 = tester[i].testStatistic(i)
succRateinPER.append(t1)
succRateinPA.append(t2)
succRateinSVM.append(t3)
print("succeeds rate: per:", t1, " pa:", t2, " svm:", t3)
if len(sys.argv) > 3: # submit mode
Testing.Testing.testerSubmit(tester, td)
# tester.test(td)
else:
plt.plot(testArgs, succRateinSVM, label="SVM {}".format(tr))
succRateinSVM = []
plt.plot(testArgs, succRateinPA, label="PA {} ".format(tr))
succRateinPA = []
"\n")
# @Kmeans
centroids = Kmeans(x_train, 11)
#print("\nCluster Centroids after Kmeans: \n", centroids)
# PCA
pcaValues = PCA(x_train, 5)
print("\nPCA with 5 features: \n")
random_forest_predictions = RandomForest(x_train, y_train, x_test)
print("Accuracy of Random Forest: ",
Accuracy(y_test, random_forest_predictions))
predicitons = SklearnSupervisedLearning(x_train, y_train, x_test)
test = Testing(predicitons, x_test, y_test, True)
test.run()
SklearnVotingClassifier(x_train, y_train, x_test)
testVoting = Testing([SklearnVotingClassifier(x_train, y_train, x_test)],
x_test, y_test, True)
testVoting.run()
endTime = timeit.default_timer()
print("Total Project time ", endTime - startTime)
tuned_parameters = [{
'kernel': ['rbf'],
'gamma': [1e-3],
'C': [1, 10, 100, 500, 1000]
}, {
'kernel': ['linear'],
# -*- coding: utf-8 -*-
from Testing import *
t = Testing()
t.init()
t.setTime("22h02m0s")
t.setRef(1, "2h31m49s", "89º15'51''", "22h04m20s", "0º27'09''", "36º49'17''")
t.setRef(2, "18h36m56s", "38º47'03''", "22h05m07s", "78º10'04''", "70º05'19''")
#t.goto1("100","160")
t.goto("18h03m48s", "24º23'00''", "22h06m52s")
#t.goto("13h17m55s", "8º29'04''", "22h07m51s")
