logai.analysis package

Submodules

logai.analysis.anomaly_detector module

class logai.analysis.anomaly_detector.AnomalyDetectionConfig(algo_name: str = 'one_class_svm', algo_params: object | None = None, custom_params: object | None = None)

Bases: Config

Config class for AnomalyDetector.

Parameters:

  • algo_name – The algorithm name.

  • algo_params – The algorithm parameters.

  • custom_params – Additional customized parameters.

algo_name: str
algo_params: object
custom_params: object
classmethod from_dict(config_dict)

Loads a config from a config dict.

Parameters:

config_dict – The config parameters in a dict.

class logai.analysis.anomaly_detector.AnomalyDetector(config: AnomalyDetectionConfig)

Bases: object

fit(log_features: DataFrame)

Trains an anomaly detection given the training dataset.

Parameters:

log_features – The training dataset.

predict(log_features: DataFrame) → DataFrame

Predicts anomalies given the test dataset.

Parameters:

log_features – The test dataset.

Returns:

A pandas dataframe containing the prediction results.

logai.analysis.clustering module

class logai.analysis.clustering.Clustering(config: ClusteringConfig)

Bases: object

Clustering Application class defines log clustering workflow. It includes which algorithm to use.

fit(log_features: DataFrame)

Fit method of Clustering algorithm, to train on the given log features data.

Parameters:

log_features – The training log features data.

predict(log_features: DataFrame) → Series

Predict method of Clustering algorithm, to run inference on given test log features.

Parameters:

log_features – The test log features data.

Returns:

The cluster output (label).

class logai.analysis.clustering.ClusteringConfig(algo_name: str = 'dbscan', algo_params: object | None = None, custom_params: object | None = None)

Bases: Config

Config class for Clustering algorithms.

Parameters:

  • algo_name – The algorithm name.

  • algo_params – The algorithm parameters.

  • custom_params – Additional customized parameters.

algo_name: str
algo_params: object
custom_params: object
classmethod from_dict(config_dict)

Loads a config from a config dict.

Parameters:

config_dict – The config parameters in a dict.

logai.analysis.nn_anomaly_detector module

class logai.analysis.nn_anomaly_detector.NNAnomalyDetector(config: AnomalyDetectionConfig)

Bases: object

fit(train_data: ForecastNNVectorizedDataset, dev_data: ForecastNNVectorizedDataset)

Trains an anomaly detection given the training and validation datasets.

Parameters:

  • train_data – The training dataset.

  • dev_data – The validation dataset

predict(test_data: ForecastNNVectorizedDataset)

Predicts anomalies given the test dataset.

Parameters:

test_data – The test dataset.

Returns:

A pandas dataframe containing the prediction results.

Module contents