Example usage:
from sklearn import datasets, ensemble
import numpy as np
from roc import roc
rf = ensemble.RandomForestClassifier()
iris = datasets.load_iris()
#changing ternary to binary problem, fitting model
adjustedTargets = list(map(lambda x: 0 if x == 0 else 1, iris.target))
rf = rf.fit(iris.data, adjustedTargets)
#nabbing labels and confidences
probas = rf.predict_proba(iris.data)
labels, class_0_confidence, class_1_confidence = adjustedTargets, probas[:,0], probas[:,0]
#roc-ing
roc(labels, class_1_confidence, 'my awesome roc', dst='roc_awesome_sample.png')