import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
%matplotlib inline
from sklearn import tree
from sklearn.externals.six import StringIO
import re

df = pd.read_csv('fruit.csv')
fruitnames = {1: 'Orange', 2: 'Pear', 3: 'Apple'}
colors = {1: '#e09028', 2: '#55aa33', 3: '#cc3333'}
fruitlist = ['Orange', 'Pear', 'Apple']

df.sort('fruit_id', inplace=True) # This is important because the factorizer assigns numbers
    # based on the order the first label is encountered, e.g. if the first instance had
    # fruit = 3, the y value would be 0.


df['is_train'] = np.random.uniform(0, 1, len(df)) <= .75 # randomly assign training and testing set
train, test = df[df['is_train']==True], df[df['is_train']==False]
features = ['color_id', 'elongatedness', 'weight', 'sweetness', 'acidity']
y, _ = pd.factorize(train['fruit_id'])
clf = tree.DecisionTreeClassifier()
clf = clf.fit(train[features], y)
preds = clf.predict(test[features])
test_result = pd.crosstab(np.array([fruitnames[x] for x in test['fruit_id']]), 
                      np.array([fruitnames[x+1] for x in preds]), rownames=['actual'], colnames=['predicted'])
test_result

# Repetition 1

df['is_train'] = np.random.uniform(0, 1, len(df)) <= .75 # randomly assign training and testing set
train, test = df[df['is_train']==True], df[df['is_train']==False]
features = ['color_id', 'elongatedness', 'weight', 'sweetness', 'acidity']
y, _ = pd.factorize(train['fruit_id'])
clf = tree.DecisionTreeClassifier()
clf = clf.fit(train[features], y)
dot_data = StringIO() 
tree.export_graphviz(clf, out_file=dot_data) 
tree_string = dot_data.getvalue()
# replace feature numbers with feature names
tree_string = re.sub('gini = 0\.[0-9]+\\\\n', '', tree_string)
for i, feature in enumerate(features):
    tree_string = re.sub('X\[{}\]'.format(i), feature, tree_string)
# repace lists of numeric label assignments with label name
for result in re.finditer('\[[ ]+([\d]+)\.[ ]+([\d]+)\.[ ]+([\d]+)\.\]', tree_string):
    nums = []
    for i in range(0,3):
        nums.append(int(result.group(i+1)))
    if nums[0] > nums[1]:
        if nums[0] > nums[2]:
            tree_string = re.sub('\[[ ]+{}\.[ ]+{}\.[ ]+{}\.\]'.format(nums[0], nums[1], nums[2]), fruitlist[0], tree_string)
        else:
            tree_string = re.sub('\[[ ]+{}\.[ ]+{}\.[ ]+{}\.\]'.format(nums[0], nums[1], nums[2]), fruitlist[2], tree_string)
    elif nums[1] > nums[2]:
        tree_string = re.sub('\[[ ]+{}\.[ ]+{}\.[ ]+{}\.\]'.format(nums[0], nums[1], nums[2]), fruitlist[1], tree_string)
    else:
        tree_string = re.sub('\[[ ]+{}\.[ ]+{}\.[ ]+{}\.\]'.format(nums[0], nums[1], nums[2]), fruitlist[2], tree_string)
with open('simple.dotfile', 'w+') as f:
    f.write(tree_string)
# normally this would be done with libraries like pydot or networkx, but
# I'm having trouble getting them to work in Python 3.4.2 under Windows,
# so I'll just call the shell executable directly
!dot.exe -Tpng simple.dotfile > simpletree.png
from IPython.core.display import Image
Image( filename ='simpletree.png')

# Repetition 2

df['is_train'] = np.random.uniform(0, 1, len(df)) <= .75 # randomly assign training and testing set
train, test = df[df['is_train']==True], df[df['is_train']==False]
features = ['color_id', 'elongatedness', 'weight', 'sweetness', 'acidity']
y, _ = pd.factorize(train['fruit_id'])
clf = tree.DecisionTreeClassifier()
clf = clf.fit(train[features], y)
dot_data = StringIO() 
tree.export_graphviz(clf, out_file=dot_data) 
tree_string = dot_data.getvalue()
# replace feature numbers with feature names
tree_string = re.sub('gini = 0\.[0-9]+\\\\n', '', tree_string)
for i, feature in enumerate(features):
    tree_string = re.sub('X\[{}\]'.format(i), feature, tree_string)
# repace lists of numeric label assignments with label name
for result in re.finditer('\[[ ]+([\d]+)\.[ ]+([\d]+)\.[ ]+([\d]+)\.\]', tree_string):
    nums = []
    for i in range(0,3):
        nums.append(int(result.group(i+1)))
    if nums[0] > nums[1]:
        if nums[0] > nums[2]:
            tree_string = re.sub('\[[ ]+{}\.[ ]+{}\.[ ]+{}\.\]'.format(nums[0], nums[1], nums[2]), fruitlist[0], tree_string)
        else:
            tree_string = re.sub('\[[ ]+{}\.[ ]+{}\.[ ]+{}\.\]'.format(nums[0], nums[1], nums[2]), fruitlist[2], tree_string)
    elif nums[1] > nums[2]:
        tree_string = re.sub('\[[ ]+{}\.[ ]+{}\.[ ]+{}\.\]'.format(nums[0], nums[1], nums[2]), fruitlist[1], tree_string)
    else:
        tree_string = re.sub('\[[ ]+{}\.[ ]+{}\.[ ]+{}\.\]'.format(nums[0], nums[1], nums[2]), fruitlist[2], tree_string)
with open('simple.dotfile', 'w+') as f:
    f.write(tree_string)
# normally this would be done with libraries like pydot or networkx, but
# I'm having trouble getting them to work in Python 3.4.2 under Windows,
# so I'll just call the shell executable directly
!dot.exe -Tpng simple.dotfile > simpletree.png
from IPython.core.display import Image
Image( filename ='simpletree.png')

# Repetition 3

df['is_train'] = np.random.uniform(0, 1, len(df)) <= .75 # randomly assign training and testing set
train, test = df[df['is_train']==True], df[df['is_train']==False]
features = ['color_id', 'elongatedness', 'weight', 'sweetness', 'acidity']
y, _ = pd.factorize(train['fruit_id'])
clf = tree.DecisionTreeClassifier()
clf = clf.fit(train[features], y)
dot_data = StringIO() 
tree.export_graphviz(clf, out_file=dot_data) 
tree_string = dot_data.getvalue()
# replace feature numbers with feature names
tree_string = re.sub('gini = 0\.[0-9]+\\\\n', '', tree_string)
for i, feature in enumerate(features):
    tree_string = re.sub('X\[{}\]'.format(i), feature, tree_string)
# repace lists of numeric label assignments with label name
for result in re.finditer('\[[ ]+([\d]+)\.[ ]+([\d]+)\.[ ]+([\d]+)\.\]', tree_string):
    nums = []
    for i in range(0,3):
        nums.append(int(result.group(i+1)))
    if nums[0] > nums[1]:
        if nums[0] > nums[2]:
            tree_string = re.sub('\[[ ]+{}\.[ ]+{}\.[ ]+{}\.\]'.format(nums[0], nums[1], nums[2]), fruitlist[0], tree_string)
        else:
            tree_string = re.sub('\[[ ]+{}\.[ ]+{}\.[ ]+{}\.\]'.format(nums[0], nums[1], nums[2]), fruitlist[2], tree_string)
    elif nums[1] > nums[2]:
        tree_string = re.sub('\[[ ]+{}\.[ ]+{}\.[ ]+{}\.\]'.format(nums[0], nums[1], nums[2]), fruitlist[1], tree_string)
    else:
        tree_string = re.sub('\[[ ]+{}\.[ ]+{}\.[ ]+{}\.\]'.format(nums[0], nums[1], nums[2]), fruitlist[2], tree_string)
with open('simple.dotfile', 'w+') as f:
    f.write(tree_string)
# normally this would be done with libraries like pydot or networkx, but
# I'm having trouble getting them to work in Python 3.4.2 under Windows,
# so I'll just call the shell executable directly
!dot.exe -Tpng simple.dotfile > simpletree.png
from IPython.core.display import Image
Image( filename ='simpletree.png')