Coverage for tadkit/base/tadlearner.py: 100%

1from typing import Sequence, Optional, Protocol, runtime_checkable 

2 

3from tadkit.base.typing import ParamsDescription, Array 

4 

5 

6@runtime_checkable 

7class TADLearner(Protocol): 

8 """Abstract class of Time Anomaly Detection Learner (model). 

9 

10 Avoid explicit inheritance from this class. Better to simply do it implicitly. 

11 

12 Methods: 

13 fit: Fit the learner on input data. 

14 score_samples: The measure of normality of an observation according to the fitted model. 

15 The lower, the more abnormal. 

16 predict: Predict if a particular sample is an outlier or not. For each observation, tells 

17 whether or not (+1 or -1) it should be considered as an inlier according to the fitted model. 

18 

19 Class attributes: 

20 params_description: Description of the arguments of the __init__ method. See examples in the catalog. 

21 required_properties: Get the properties that the input data must satisfies. See examples in the catalog. 

22 

23 Example: 

24 >>> assert isinstance(MyLearner, TADLearner) 

25 >>> MyLearner.required_properties # The required property of input data 

26 >>> MyLearner.params_description # The description of the params 

27 >>> params = ... # Params to initiate learner 

28 >>> learner = MyLearner(**params) 

29 >>> learner.fit(X) # X, y must satisfy MyLearner.required_properties 

30 >>> score_sample_pred = learner.score_samples(X_test) 

31 """ 

32 

33 required_properties: Sequence[str] = [] 

34 params_description: ParamsDescription = {} 

35 

36 def fit(self, X: Array, y: Optional[Array] = None) -> "TADLearner": ... 

37 

38 def score_samples(self, X: Array) -> Array: 

39 """ 

40 The measure of normality of an observation according to the fitted model. 

41 Scikit-learn compatible. 

42 

43 Parameters 

44 ---------- 

45 X : {array-like, sparse matrix} of shape (n_samples, n_features) 

46 The input samples. 

47 

48 Returns 

49 ------- 

50 scores : ndarray of shape (n_samples,) 

51 The anomaly score of the input samples. 

52 The lower, the more abnormal. 

53 """ 

54 ... 

55 

56 def predict(self, X: Array) -> Array: 

57 """ 

58 Predict if a particular sample is an outlier or not. 

59 Scikit-learn compatible. 

60 

61 Parameters 

62 ---------- 

63 X : {array-like, sparse matrix} of shape (n_samples, n_features) 

64 The input samples. 

65 

66 Returns 

67 ------- 

68 is_inlier : ndarray of shape (n_samples,) 

69 For each observation, tells whether or not (+1 or -1) it should 

70 be considered as an inlier according to the fitted model. 

71 """ 

72 ...