ativ_1.py

import numpy as np

from sklearn.neighbors import KNeighborsClassifier
from sklearn.naive_bayes import GaussianNB
from sklearn.tree import DecisionTreeClassifier
from numpy import random

X = np.genfromtxt('bases/iris.data', dtype=None, delimiter=',', usecols=(0, 1, 2, 3))
y = np.genfromtxt('bases/iris.data', dtype=None, delimiter=',', usecols=[4])

values_for_training = random.choice(range(0, X.shape[0]), X.shape[0]/4)

values_for_test = np.delete(range(0, X.shape[0]), values_for_training)

X_training = [X[i] for i in values_for_training]
y_training = [y[i] for i in values_for_training]

X_test = [X[i] for i in values_for_test]
y_test = [y[i] for i in values_for_test]

# eh possivel fazer isso pois todas as classes herdao de ClassifierMixin
configurations = [
    ("Knn1", KNeighborsClassifier, 1),
    ("Knn3", KNeighborsClassifier, 3),
    ("Knn5", KNeighborsClassifier, 5),
    ("Naive bayes", GaussianNB, None),
    ("Decision tree | entropy criterion", DecisionTreeClassifier, 'entropy'),
    ("Decision tree | gini criterion", DecisionTreeClassifier, 'gini'),
]

classifiers = []
for classifier in configurations:
    classifiers.append((classifier[1](classifier[2]), classifier[0]))

for classifier in classifiers:
    classifier[0].fit(X_training, y_training)

for classifier in classifiers:
    print '{} score is: {}'.format(classifier[1], classifier[0].score(X_test, y_test))