import monkdata as m

import dtree as d

import numpy as np

import matplotlib.pyplot as plt

import drawtree_qt5 as dt5

import random

import statistics

#assignment 1

entropy1 = d.entropy(m.monk1)

entropy2 = d.entropy(m.monk2)

entropy3 = d.entropy(m.monk3)

print('entropy monk1='+str(entropy1))

print('entropy monk2='+str(entropy2))

print('entropy monk3='+str(entropy3))

#assignment 2

N_points = 1000

n_bins = 20

#plt.figure(1)

fig, axs = plt.subplots(1, 2, tight_layout=True)

#high entropy

s = np.random.uniform(0, 1, N_points)

h = axs[0].hist(s, bins=n_bins)

entropy = 0

for v in h[0]:

if v == 0:

continue

en1 = -v/N_points * d.log2(v/N_points)

entropy = entropy + en1

axs[0].set_title("entropy="+str(entropy))

print('entropy for uniform distribution random sampling='+str(entropy))

#low entropy

s = np.random.randn(N_points)

h = axs[1].hist(s, bins=n_bins)

entropy = 0

for v in h[0]:

if v == 0:

continue

en1 = -v/N_points * d.log2(v/N_points)

entropy = entropy + en1

axs[1].set_title("entropy="+str(entropy))

print('entropy for normal distribution random sampling='+str(entropy))

#assignment 3

aveGain = {'monk1': {}, 'monk2': {}, 'monk3': {}}

i = 1

j = 1

for monk in [m.monk1, m.monk2, m.monk3]:

for attribute in m.attributes:

monkN = 'monk' + str(i)

attributeN = 'a' + str(j)

attrN = aveGain[monkN]

attrN[attributeN] = d.averageGain(monk, attribute)

aveGain[monkN] = attrN

j = j+1

i = i+1

j = 1

print(aveGain)

#assignment 5

#monk1: a5 is selected

monk1_1 = d.select(m.monk1, m.attributes[4], 1)

monk1_2 = d.select(m.monk1, m.attributes[4], 2)

monk1_3 = d.select(m.monk1, m.attributes[4], 3)

monk1_4 = d.select(m.monk1, m.attributes[4], 4)

aveGain1 = {'monk1_1': {}, 'monk1_2': {}, 'monk1_3': {}, 'monk1_4': {}}

i = 1

j = 1

for monk in [monk1_1, monk1_2, monk1_3, monk1_4]:

for attribute in m.attributes:

monkN = 'monk1_' + str(i)

attributeN = 'a' + str(j)

attrN = aveGain1[monkN]

attrN[attributeN] = d.averageGain(monk, attribute)

aveGain1[monkN] = attrN

j = j+1

i = i+1

j = 1

print(aveGain1)

#branch 1: N/A

#branch 2: a4

#branch 3: a6

#branch 4: a1

monkData = {'monk1_2': {'data': monk1_2, 'branch': 3}, 'monk1_3': {'data': monk1_3, 'branch': 5}, 'monk1_4': {'data':

monk1_4, 'branch': 0}}

monk1_maj = d.mostCommon(m.monk1)

monk1_1_maj = d.mostCommon(monk1_1)

print('majority class for monk1: ' + str(monk1_maj))

print('majority class for monk1_1: ' + str(monk1_1_maj))

for mo in monkData:

moName = mo

attrNo = monkData[mo]['branch']

monk_maj = d.mostCommon(monkData[mo]['data'])

print('majority class for ' + mo + ': ' + str(monk_maj))

for attrVal in m.attributes[attrNo].values:

monkTmp = d.select(monkData[mo]['data'], m.attributes[attrNo], attrVal)

monk_maj = d.mostCommon(monkTmp)

print('majority class for ' + mo + ', partition ' + str(attrVal) + ': ' + str(monk_maj))

#tree1 = d.buildTree(m.monk1, m.attributes, 2)

#dt5.drawTree(tree1)

t1 = d.buildTree(m.monk1, m.attributes)

print("Training data Monk1", 1-d.check(t1, m.monk1))

print("Testing data Monk1", 1-d.check(t1, m.monk1test))

t2 = d.buildTree(m.monk2, m.attributes)

print("Training data Monk2", 1-d.check(t2, m.monk2))

print("Testing data Monk2", 1-d.check(t2, m.monk2test))

t3 = d.buildTree(m.monk3, m.attributes)

print("Training data Monk3", 1-d.check(t3, m.monk3))

print("Testing data Monk3", 1-d.check(t3, m.monk3test))

#tree2 = d.buildTree(m.monk1, m.attributes)

#dt5.drawTree(tree2)

#assignment6

def partition(data, fraction):

return ldata[:breakPoint], ldata[breakPoint:]

def pruneTree(tree, oldError, validationTree) :

return pruneTree(prunedTrees[index_of_bE], bestError, validationTree)

# Initialize lists of trees to be pruned

fractions = list((0.3, 0.4, 0.5, 0.6, 0.7, 0.8))

monk1Trees = list()

for i in range(0, len(fractions)) :

monk1Trees.append(list())

monk1ValTrees = list()

for i in range(len(fractions)) :

monk1ValTrees.append(list())

monk3Trees = list()

for i in range(len(fractions)) :

monk3Trees.append(list())

monk3ValTrees = list()

for i in range(len(fractions)) :

monk3ValTrees.append(list())

# Randomize training and validation sets

for i in range(len(fractions)) :

for j in range(128) :

monk1train, monk1val = partition(m.monk1, fractions[i])

monk3train, monk3val = partition(m.monk3, fractions[i])

monk1Trees[i].append(d.buildTree(monk1train, m.attributes))

monk1ValTrees[i].append(monk1val)

monk3Trees[i].append(d.buildTree(monk3train, m.attributes))

monk3ValTrees[i].append(monk3val)

Array_1 = np.zeros((6,128))

Array_3 = np.zeros((6,128))

txt = open("as6_data.txt", "w")

txt.write("monk1\n")

for i in range(len(monk1Trees)) :

txt.write(str("\n\n" + "Fraction: " + str(fractions[i]) + "\n\n"))

for j in range(len(monk1Trees[i])) :

prunedTree, prunedError = pruneTree(monk1Trees[i][j], d.check(monk1Trees[i][j], monk1ValTrees[i][j]), monk1ValTrees[i][j])

Array_1[i][j] = prunedError

txt.write(str(prunedError) + "\n")

txt.write("Finished monk1")

txt.write("\n\n")

txt.write("monk3\n")

for i in range(len(monk3Trees)) :

txt.write(str("\n\n" + "Fraction: " + str(fractions[i]) + "\n\n"))

for j in range(len(monk3Trees[i])) :

prunedTree, prunedError = pruneTree(monk3Trees[i][j], d.check(monk3Trees[i][j], monk3ValTrees[i][j]), monk3ValTrees[i][j])

Array_3[i][j] = prunedError

txt.write(str(prunedError) + "\n")

txt.write("Finished monk3")

x1 = [0.3, 0.4, 0.5, 0.6, 0.7, 0.8]

y1 = list()

for i in range(6):

y1.append(np.mean(Array_1[i]))

plot1 = plt.figure(2)

plt.plot(x1, y1)

plt.title('Mean of validation performance on pruned tree, monk1')

plt.xlabel('Fraction of Training Data')

plt.ylabel('Validation Data Mean Performance')

x2 = [0.3, 0.4, 0.5, 0.6, 0.7, 0.8]

y2 = list()

for i in range(6):

y2.append(statistics.variance(Array_1[i]))

plot2 = plt.figure(3)

plt.plot(x2, y2)

plt.title('Variance of validation performance on pruned tree, monk1')

plt.xlabel('Fraction of Training Data')

plt.ylabel('Validation Data Performance Variance')

x3 = [0.3, 0.4, 0.5, 0.6, 0.7, 0.8]

y3 = list()

for i in range(6):

y3.append(np.mean(Array_3[i]))

plot3 = plt.figure(4)

plt.plot(x3, y3)

plt.title('Mean of validation performance on pruned tree, monk3')

plt.xlabel('Fraction of Training Data')

plt.ylabel('Validation Data Mean Performance')

x4 = [0.3, 0.4, 0.5, 0.6, 0.7, 0.8]

y4 = list()

for i in range(6):

y4.append(statistics.variance(Array_3[i]))

plot4 = plt.figure(5)

plt.plot(x4, y4)

plt.title('Variance of validation performance on pruned tree, monk3')

plt.xlabel('Fraction of Training Data')

plt.ylabel('Validation Data Performance Variance')

plt.show()