분류 모델에 대한 평가 지표를 출력하는 코드
Example1
from classification_report import ClassificationReport
from pprint import pprint
y_true = [0, 1, 2, 2, 2]
y_pred = [0, 0, 2, 2, 1]
target_names = ["class 0", "class 1", "class 2"]
matrix = ClassificationReport(class_name=target_names)
matrix.compute(y_true, y_pred)
print(matrix.print_report())
pprint(matrix.report_dict()) precision recall f1_score n_sample
class 0 0.500 1.000 0.667 1
class 1 0.000 0.000 0.000 1
class 2 1.000 0.667 0.800 3
------------------------------------------------
accuracy 0.000 0.000 0.600 5
{'accuracy': 0.6,
'class 0': {'f1_score': 0.6666666657777777,
'n_sample': 1,
'precision': 0.49999999975,
'recall': 0.9999999989999999},
'class 1': {'f1_score': 0.0, 'n_sample': 1, 'precision': 0.0, 'recall': 0.0},
'class 2': {'f1_score': 0.7999999992,
'n_sample': 3,
'precision': 0.9999999995,
'recall': 0.6666666664444444},
'total_sample': 5}Example2
from classification_report import ClassificationReport
y_true = [0, 1, 2, 2, 2]
y_pred = [0, 0, 2, 2, 1]
target_names = ["class 0", "class 1", "class 2"]
matrix = ClassificationReport(class_name=target_names)
matrix.compute(y_true, y_pred)
print(matrix.print_report())
y_true2 = [2, 2, 1, 0, 0]
y_pred2 = [2, 1, 1, 0, 0]
matrix.compute(y_true2, y_pred2)
print(matrix.print_report()) precision recall f1_score n_sample
class 0 0.500 1.000 0.667 1
class 1 0.000 0.000 0.000 1
class 2 1.000 0.667 0.800 3
------------------------------------------------
accuracy 0.000 0.000 0.600 5
precision recall f1_score n_sample
class 0 0.750 1.000 0.857 3
class 1 0.333 0.500 0.400 2
class 2 1.000 0.600 0.750 5
------------------------------------------------
accuracy 0.000 0.000 0.700 10Example3
from classification_report import ClassificationReport
y_true = [0, 1, 2, 2, 2]
y_pred = [0, 0, 2, 2, 1]
target_names = ["class 0", "class 1", "class 2"]
matrix = ClassificationReport(class_name=target_names)
matrix.compute(y_true, y_pred)
print(matrix.print_report())
matrix.reset()
y_true2 = [2, 2, 1, 0, 0]
y_pred2 = [2, 1, 1, 0, 0]
matrix.compute(y_true2, y_pred2)
print(matrix.print_report()) precision recall f1_score n_sample
class 0 0.500 1.000 0.667 1
class 1 0.000 0.000 0.000 1
class 2 1.000 0.667 0.800 3
------------------------------------------------
accuracy 0.000 0.000 0.600 5
precision recall f1_score n_sample
class 0 1.000 1.000 1.000 2
class 1 0.500 1.000 0.667 1
class 2 1.000 0.500 0.667 2
------------------------------------------------
accuracy 0.000 0.000 0.800 5
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent