SCOPE/evaluate.py at main · PhilJahn/SCOPE · 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
import river.metrics as rmetrics
from clustpy.metrics import unsupervised_clustering_accuracy, PairCountingScores
import numpy as np
import logging

logger = logging.getLogger(__name__)

class Purity(rmetrics.base.MultiClassMetric):

	@property
	def works_with_weights(self):
		return False

	def get(self):
		purity = 0
		num = 0
		#print(self.cm)
		for j in self.cm.classes:  # predict
			maxi = 0
			for i in self.cm.classes:  # label
				num += self.cm[i][j]
				if maxi < self.cm[i][j]:
					maxi = self.cm[i][j]
			purity += maxi

		return purity / num

def getMetrics(labels, predictions):
	stream_ari = rmetrics.AdjustedRand()
	stream_nmi = rmetrics.NormalizedMutualInfo()
	stream_ami = rmetrics.AdjustedMutualInfo()
	stream_completeness = rmetrics.Completeness()
	stream_fowl = rmetrics.FowlkesMallows()
	stream_homogeneity = rmetrics.Homogeneity()
	stream_purity = Purity()

	for i in range(len(predictions)):
		pred = predictions[i]
		label = labels[i]
		stream_ari.update(label, pred)
		stream_nmi.update(label, pred)
		stream_ami.update(label, pred)
		stream_completeness.update(label, pred)
		stream_fowl.update(label, pred)
		stream_homogeneity.update(label, pred)
		stream_purity.update(label, pred)

	ari = stream_ari.get()
	nmi = stream_nmi.get()
	ami = stream_ami.get()
	completeness = stream_completeness.get()
	fowl = stream_fowl.get()
	homogeneity = stream_homogeneity.get()
	purity = stream_purity.get()
	acc = unsupervised_clustering_accuracy(np.array(labels), np.array(predictions))
	paircounting = PairCountingScores(np.array(labels), np.array(predictions))
	f1 = paircounting.f1()
	prec = paircounting.precision()
	recall = paircounting.recall()

	clunum = len(np.unique(predictions))
	trueclunum = len(np.unique(labels))

	cm = stream_purity.cm

	metrics = {"accuracy": acc, "ARI": ari, "NMI": nmi, "AMI": ami, "completeness": completeness, "fowl": fowl,
			   "homogeneity": homogeneity, "purity": purity, "F1": f1, "precision": prec, "recall": recall,
			   "cluster_num": clunum}
	return metrics, cm

def printMetrics(labels, predictions):
	metrics, cm = getMetrics(labels, predictions)
	print(cm)
	print(metrics)