add anomaly detection algorithms

[pep8speaks]

Thanks for update, @Lopa10ko!

• In the file examples/automl_example/api_example/time_series/ts_anomaly_detection/custom_liman_example.py:

Line 76:19: F541 f-string is missing placeholders Line 78:19: F541 f-string is missing placeholders

• In the file fedot_ind/api/main.py:

Line 212:9: F841 local variable 'fit_function' is assigned to but never used Line 485:17: F541 f-string is missing placeholders Line 487:17: F541 f-string is missing placeholders

• In the file fedot_ind/api/utils/industrial_strategy.py:

Line 123:30: F541 f-string is missing placeholders Line 175:32: F841 local variable 'n_classes' is assigned to but never used

• In the file fedot_ind/core/architecture/pipelines/abstract_pipeline.py:

Line 193:121: E501 line too long (121 > 120 characters) Line 194:121: E501 line too long (121 > 120 characters)

• In the file fedot_ind/core/ensemble/kernel_ensemble.py:

Line 71:74: W503 line break before binary operator

• In the file fedot_ind/core/metrics/anomaly_detection/function.py:

Line 65:121: E501 line too long (122 > 120 characters) Line 337:9: F841 local variable 'scale_func' is assigned to but never used

• In the file fedot_ind/core/metrics/metrics_implementation.py:

Line 17:1: F811 redefinition of unused 'np' from line 4 Line 93:73: W503 line break before binary operator Line 515:28: W503 line break before binary operator

• In the file fedot_ind/core/models/detection/anomaly/algorithms/lstm_autoencoder_detector.py:

Line 87:22: F841 local variable 'losses' is assigned to but never used

• In the file fedot_ind/core/models/nn/network_modules/layers/pooling_layers.py:

Line 4:1: F403 'from fastai.layers import *' used; unable to detect undefined names Line 103:21: F405 'AdaptiveConcatPool1d' may be undefined, or defined from star imports: fastai.layers Line 155:40: F405 'LinBnDrop' may be undefined, or defined from star imports: fastai.layers

• In the file fedot_ind/core/models/ts_forecasting/ssa_forecaster.py:

Line 89:121: E501 line too long (128 > 120 characters)

• In the file fedot_ind/core/operation/interfaces/industrial_preprocessing_strategy.py:

Line 46:121: E501 line too long (121 > 120 characters)

• In the file fedot_ind/core/repository/constanst_repository.py:

Line 18:121: E501 line too long (121 > 120 characters)

• In the file fedot_ind/core/repository/industrial_implementations/optimisation.py:

Line 554:5: F841 local variable 'other' is assigned to but never used

• In the file fedot_ind/core/repository/model_repository.py:

Line 14:1: F403 'from fedot.core.operations.evaluation.operation_implementations.data_operations.sklearn_transformations import *' used; unable to detect undefined names Line 15:121: E501 line too long (127 > 120 characters) Line 77:36: F405 'PCAImplementation' may be undefined, or defined from star imports: fedot.core.operations.evaluation.operation_implementations.data_operations.sklearn_transformations Line 78:41: F405 'FastICAImplementation' may be undefined, or defined from star imports: fedot.core.operations.evaluation.operation_implementations.data_operations.sklearn_transformations Line 79:46: F405 'PolyFeaturesImplementation' may be undefined, or defined from star imports: fedot.core.operations.evaluation.operation_implementations.data_operations.sklearn_transformations Line 144:20: F405 'ScalingImplementation' may be undefined, or defined from star imports: fedot.core.operations.evaluation.operation_implementations.data_operations.sklearn_transformations Line 145:26: F405 'NormalizationImplementation' may be undefined, or defined from star imports: fedot.core.operations.evaluation.operation_implementations.data_operations.sklearn_transformations Line 147:30: F405 'ImputationImplementation' may be undefined, or defined from star imports: fedot.core.operations.evaluation.operation_implementations.data_operations.sklearn_transformations Line 149:23: F405 'KernelPCAImplementation' may be undefined, or defined from star imports: fedot.core.operations.evaluation.operation_implementations.data_operations.sklearn_transformations

• In the file fedot_ind/tools/loader.py:

Line 74:9: F841 local variable 'folder_dict' is assigned to but never used Line 391:25: F841 local variable 'class_label_list' is assigned to but never used Line 576:53: W503 line break before binary operator Line 587:53: W503 line break before binary operator Line 627:53: W503 line break before binary operator

Comment last updated at 2024-06-04 13:04:39 UTC

[v1docq]

работа проделана колоссальная, это круто

но я бы отметил следующее:

1. Примеры лучше оформить в виде ноутбуков. Иначе к ним нельзя даже никого отослать, чтобы потенциальный пользователь изучил принцип работы.
2. Мы в примерах используем industrial_forecasting_modelling_loop, industrial_common_modelling_loop. Если предполагается, что кто-то будет этим пользоваться, типа для упрощения проведения экспериментов, то я бы поработал над нэйминогом. Если это один из основных инструментов, то его импорт – неудобен
3. Классно, что добавилось столько алго для детектирования. Будет ещё круче, если к каждому из них добавится по самому простому юниту. На примере PR по DeepAR, можно понять, зачем это делать
4. Надо проверить работу созданных примеров: в одном случае валятся ошибки, в другом – нет исходных данных

1. Можно на стажеров оформить таску.
2. Подумаю как это отрефакторить
3. В процессе
4. Правим