Python_ID3/main.py at master · juliamarcinkowska/Python_ID3 · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
import csv
import math
import random
import statistics

from animal import Animal
from node import build


def read_instances(file):  # create list of Animal objects from csv file
    with open(file, newline='') as csvfile:
        reader = csv.reader(csvfile, delimiter=',')
        objects = [Animal(row) for row in reader]
    return objects


def print_tree(node, counter, text):  # print tree with indentation (recursively) based on root node
    if not node.branches:  # leaves
        text += ' ' + node.name
        print(text)
        return
    print(text)
    for n in range(len(node.branches)):  # branches
        text = node.name + " = " + str(node.branches[n].value) + ':'
        if node.branches:
            print_tree(node.branches[n], counter + 1, text.rjust(counter * 2 + len(text)))


def classify(node, animal):  # recursive function that classifies animal based on nodes
    if node.branches:
        for b in node.branches:
            if b.value == animal.__getattribute__(node.name):
                return classify(b, animal)
    else:
        if node.name == animal.animal_type:
            return True
        else:  # it prints only incorrectly classified animals for legibility
            print(animal.animal_name + " is " + animal.animal_type + " and is classified as " + node.name)
            return False


def main():
    attributes = [["hair", [0, 1]],
                  ["feathers", [0, 1]],
                  ["eggs", [0, 1]],
                  ["milk", [0, 1]],
                  ["airborne", [0, 1]],
                  ["aquatic", [0, 1]],
                  ["predator", [0, 1]],
                  ["toothed", [0, 1]],
                  ["backbone", [0, 1]],
                  ["breathes", [0, 1]],
                  ["venomous", [0, 1]],
                  ["fins", [0, 1]],
                  ["legs", [0, 1, 2, 3, 4, 5, 6, 7, 8, 9]],
                  ["tail", [0, 1]],
                  ["domestic", [0, 1]],
                  ["catsize", [0, 1]],
                  ["animal_type", ["mammal", "bird", "reptile", "fish", "amphibian", "insect", "invertebrate"]]]
    classes = attributes[16][1]
    attributes.pop(16)  # copy animal_type to classes and delete from attributes

    instances = read_instances("zoo.csv")
    root_node = build(instances, attributes, classes, None, "mammal", [])
    print_tree(root_node, 0, '')  # print tree built with all instances

    final_results = []
    for j in range(10):
        results = 0  # testing algorithm
        print("  ------  j = " + str(j) + "  ------  ")
        for i in range(10):
            success = 0
            random.shuffle(instances)  # shuffling of all instances
            test_instances = instances.copy()
            test_set = test_instances[:math.ceil(len(test_instances) / 10)]  # 10% of instances are test set
            training_set = test_instances[math.ceil(len(test_instances) / 10):]  # 90% of instances are training set
            new_root_node = build(training_set, attributes, classes, None, "mammal",
                                  [])  # tree based on training instances
            for ts in test_set:
                if classify(new_root_node, ts):
                    success += 1
            results += success / len(test_set)
        precision = results / 10
        final_results.append(precision)
    print(final_results)
    print("Avg of final results: " + str(sum(final_results) / len(final_results)))
    print("Std of final results: " + str(statistics.stdev(final_results)))


if __name__ == "__main__":
    main()