You can not select more than 25 topics Topics must start with a chinese character,a letter or number, can include dashes ('-') and can be up to 35 characters long.

test_PyodOCSVM.py 4.2 kB

first commit Former-commit-id: 08bc23ba02cffbce3cf63962390a65459a132e48 [formerly 0795edd4834b9b7dc66db8d10d4cbaf42bbf82cb] [formerly b5010b42541add7e2ea2578bf2da537efc457757 [formerly a7ca09c2c34c4fc8b3d8e01fcfa08eeeb2cae99d]] [formerly 615058473a2177ca5b89e9edbb797f4c2a59c7e5 [formerly 743d8dfc6843c4c205051a8ab309fbb2116c895e] [formerly bb0ea98b1e14154ef464e2f7a16738705894e54b [formerly 960a69da74b81ef8093820e003f2d6c59a34974c]]] [formerly 2fa3be52c1b44665bc81a7cc7d4cea4bbf0d91d5 [formerly 2054589f0898627e0a17132fd9d4cc78efc91867] [formerly 3b53730e8a895e803dfdd6ca72bc05e17a4164c1 [formerly 8a2fa8ab7baf6686d21af1f322df46fd58c60e69]] [formerly 87d1e3a07a19d03c7d7c94d93ab4fa9f58dada7c [formerly f331916385a5afac1234854ee8d7f160f34b668f] [formerly 69fb3c78a483343f5071da4f7e2891b83a49dd18 [formerly 386086f05aa9487f65bce2ee54438acbdce57650]]]] Former-commit-id: a00aed8c934a6460c4d9ac902b9a74a3d6864697 [formerly 26fdeca29c2f07916d837883983ca2982056c78e] [formerly 0e3170d41a2f99ecf5c918183d361d4399d793bf [formerly 3c12ad4c88ac5192e0f5606ac0d88dd5bf8602dc]] [formerly d5894f84f2fd2e77a6913efdc5ae388cf1be0495 [formerly ad3e7bc670ff92c992730d29c9d3aa1598d844e8] [formerly 69fb3c78a483343f5071da4f7e2891b83a49dd18]] Former-commit-id: 3c19c9fae64f6106415fbc948a4dc613b9ee12f8 [formerly 467ddc0549c74bb007e8f01773bb6dc9103b417d] [formerly 5fa518345d958e2760e443b366883295de6d991c [formerly 3530e130b9fdb7280f638dbc2e785d2165ba82aa]] Former-commit-id: 9f5d473d42a435ec0d60149939d09be1acc25d92 [formerly be0b25c4ec2cde052a041baf0e11f774a158105d] Former-commit-id: 9eca71cb73ba9edccd70ac06a3b636b8d4093b04
5 years ago
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103
  1. import unittest
  2. from d3m import container, utils
  3. from d3m.metadata import base as metadata_base
  4. from d3m.container import DataFrame as d3m_dataframe
  5. from detection_algorithm.PyodOCSVM import OCSVMPrimitive
  6. from pyod.utils.data import generate_data
  7. from detection_algorithm.core.UODCommonTest import UODCommonTest
  8. import numpy as np
  9. class PyodOCSVMTestCase(unittest.TestCase):
  10. def setUp(self):
  11. self.maxDiff = None
  12. self.n_train = 200
  13. self.n_test = 100
  14. self.contamination = 0.1
  15. self.roc_floor = 0.8
  16. self.X_train, self.y_train, self.X_test, self.y_test = generate_data(
  17. n_train=self.n_train, n_test=self.n_test,
  18. contamination=self.contamination, random_state=42)
  19. self.X_train = d3m_dataframe(self.X_train, generate_metadata=True)
  20. self.X_test = d3m_dataframe(self.X_test, generate_metadata=True)
  21. hyperparams_default = OCSVMPrimitive.metadata.get_hyperparams().defaults()
  22. hyperparams = hyperparams_default.replace({'contamination': self.contamination, })
  23. hyperparams = hyperparams.replace({'return_subseq_inds': True, })
  24. self.primitive = OCSVMPrimitive(hyperparams=hyperparams)
  25. self.primitive.set_training_data(inputs=self.X_train)
  26. self.primitive.fit()
  27. self.prediction_labels = self.primitive.produce(inputs=self.X_test).value
  28. self.prediction_score = self.primitive.produce_score(inputs=self.X_test).value
  29. self.uodbase_test = UODCommonTest(model=self.primitive._clf,
  30. X_train=self.X_train,
  31. y_train=self.y_train,
  32. X_test=self.X_test,
  33. y_test=self.y_test,
  34. roc_floor=self.roc_floor,
  35. )
  36. def test_detector(self):
  37. self.uodbase_test.test_detector()
  38. def test_metadata(self):
  39. # print(self.prediction_labels.metadata.to_internal_simple_structure())
  40. self.assertEqual(utils.to_json_structure(self.prediction_labels.metadata.to_internal_simple_structure()), [{
  41. 'selector': [],
  42. 'metadata': {
  43. # 'top_level': 'main',
  44. 'schema': metadata_base.CONTAINER_SCHEMA_VERSION,
  45. 'structural_type': 'd3m.container.pandas.DataFrame',
  46. 'semantic_types': ['https://metadata.datadrivendiscovery.org/types/Table'],
  47. 'dimension': {
  48. 'name': 'rows',
  49. 'semantic_types': ['https://metadata.datadrivendiscovery.org/types/TabularRow'],
  50. 'length': 100,
  51. },
  52. },
  53. }, {
  54. 'selector': ['__ALL_ELEMENTS__'],
  55. 'metadata': {
  56. 'dimension': {
  57. 'name': 'columns',
  58. 'semantic_types': ['https://metadata.datadrivendiscovery.org/types/TabularColumn'],
  59. 'length': 3,
  60. },
  61. },
  62. }, {
  63. 'selector': ['__ALL_ELEMENTS__', 0],
  64. 'metadata': {
  65. 'name': 'TODS.anomaly_detection_primitives.OCSVMPrimitive0_0',
  66. 'semantic_types': ['https://metadata.datadrivendiscovery.org/types/Attribute'],
  67. 'structural_type': 'numpy.int64',
  68. },
  69. }, {
  70. 'selector': ['__ALL_ELEMENTS__', 1],
  71. 'metadata': {
  72. 'name': 'TODS.anomaly_detection_primitives.OCSVMPrimitive0_1',
  73. 'semantic_types': ['https://metadata.datadrivendiscovery.org/types/Attribute'],
  74. 'structural_type': 'numpy.int64',
  75. },
  76. }, {
  77. 'selector': ['__ALL_ELEMENTS__', 2],
  78. 'metadata': {
  79. 'name': 'TODS.anomaly_detection_primitives.OCSVMPrimitive0_2',
  80. 'semantic_types': ['https://metadata.datadrivendiscovery.org/types/Attribute'],
  81. 'structural_type': 'numpy.int64',
  82. },
  83. }])
  84. def test_params(self):
  85. params = self.primitive.get_params()
  86. self.primitive.set_params(params=params)
  87. if __name__ == '__main__':
  88. unittest.main()

全栈的自动化机器学习系统,主要针对多变量时间序列数据的异常检测。TODS提供了详尽的用于构建基于机器学习的异常检测系统的模块,它们包括:数据处理(data processing),时间序列处理( time series processing),特征分析(feature analysis),检测算法(detection algorithms),和强化模块( reinforcement module)。这些模块所提供的功能包括常见的数据预处理、时间序列数据的平滑或变换,从时域或频域中抽取特征、多种多样的检测算