y_pred = model.predict(x_test) # return reconstruction errors
theta, f1 = find_optimal_threshold(y_test, y_pred)
y_pred = classify(y_pred, theta)
metrics_report(y_test, y_pred)
scores.append(
create_experiment_report(get_metrics(y_test, y_pred),
{'window': w}))
create_checkpoint(
{'experiments': scores},
'../../models/TCN-cropped-window-embeddings-HDFS1.json')
return {'experiments': scores}
if __name__ == '__main__':
X_train = load_pickle_file(
'../../data/processed/HDFS1/X-train-HDFS1-cv1-1-block.pickle')
X_val = load_pickle_file(
'../../data/processed/HDFS1/X-val-HDFS1-cv1-1-block.pickle')
y_train = np.load(
'../../data/processed/HDFS1/y-train-HDFS1-cv1-1-block.npy')
y_val = np.load('../../data/processed/HDFS1/y-val-HDFS1-cv1-1-block.npy')
# results = train_window(X_train, X_val, y_train, y_val)
# save_experiment(results, '../../models/TCN-cropped-window-embeddings-HDFS1.json')
# results = train_tcnn(X_train, X_val, y_train, y_val)
# save_experiment(results, EXPERIMENT_PATH)
# results = train_cnn1d(X_train, X_val, y_train, y_val)
# save_experiment(results, EXPERIMENT_PATH)
if isinstance(model, LocalOutlierFactor):
y_pred = model.fit_predict(x_test) # return labels
else:
y_pred = model.predict(x_test) # return labels
y_pred = convert_predictions(y_pred)
auc_score = roc_auc_score(y_test, y_pred)
metrics_report(y_test, y_pred)
return create_report(
model_config, {
**get_metrics(y_test, y_pred), 'auc_score': float(auc_score)
})
if __name__ == '__main__':
X_train = load_pickle_file(
'../../data/processed/HDFS1/X-train-HDFS1-cv1-1-block.pickle')
X_test = load_pickle_file(
'../../data/processed/HDFS1/X-test-HDFS1-block.pickle')
y_test = np.load('../../data/processed/HDFS1/y-test-HDFS1-block.npy')
experiment_reports = {}
results = evaluate_tcnn(X_train, X_test, y_test)
experiment_reports['TCN model'] = results
print('TCN model:', json.dumps(results, indent=4, sort_keys=True))
results = evaluate_cnn1d(X_train, X_test, y_test)
experiment_reports['CNN1D model'] = results
print('CNN1D model:', json.dumps(results, indent=4, sort_keys=True))
results = evaluate_cnn2d(X_train, X_test, y_test)
"layers": [
[142],
1246,
[
100 # output_shape == input_shape
]
],
"learning_rate": 0.0016378937069540646,
"window": 45
},
# not used currently
"model_path": "../../models/aetcn/4f5f4682-1ca5-400a-a340-6243716690c0.pt",
"threshold": 0.00331703620031476
}
X = load_pickle_file(
'../../data/processed/HDFS1/X-val-HDFS1-cv1-1-block.npy')[:1000]
y = np.load('../../data/processed/HDFS1/y-val-HDFS1-cv1-1-block.npy')[:1000]
# list of matrices, a matrix == blk_id -> NumPy [[005515, ...], ...] (n_logs x 100)
# F1 = (2 * r * p) / (r + p)
n_examples = 700
sc = CustomMinMaxScaler() # range 0 -- 1
x_train = sc.fit_transform(X[:n_examples])
y_train = y[:n_examples]
x_test = sc.transform(X[n_examples:])
y_test = y[n_examples:]
model = AETCN()
for conf in zip(*params.values()):
kwargs = {k: val for k, val in zip(params.keys(), conf)}
model.set_params(**kwargs)
print(f'Model current hyperparameters are: {kwargs}.')
model.fit(x_train)
y_pred = model.predict(x_test) # return reconstruction errors
theta, f1 = find_optimal_threshold(y_test, y_pred)
y_pred = classify(y_pred, theta)
metrics_report(y_test, y_pred)
scores.append(
create_experiment_report(get_metrics(y_test, y_pred), kwargs))
create_checkpoint({'experiments': scores}, EXPERIMENT_PATH)
return {'experiments': scores}
if __name__ == '__main__':
X_train = load_pickle_file(
'../../data/interim/HDFS1/train-data-Drain3-HDFS1-cv1-1.binlog')
y_train = load_labels(
'../../data/interim/HDFS1/train-labels-HDFS1-cv1-1.csv')
X_val = load_pickle_file(
'../../data/interim/HDFS1/val-data-Drain3-HDFS1-cv1-1.binlog')
y_val = load_labels('../../data/interim/HDFS1/val-labels-HDFS1-cv1-1.csv')
results = train_autoencoder(X_train, X_val, y_train, y_val)
save_experiment(results, EXPERIMENT_PATH)