@@ -409,12 +409,20 @@ y = np.random.randint(1, 100, size=(1000, 1))
409409
410410data_model = RegressionDataModel(X, y)
411411
412+ ```
413+
414+ directly write an attribute to the data model
415+ ``` python
412416data_model._update_attributes(' X_shape' , (1000 , 10 , 10 ))
413417data_model.X_shape
418+ >> > (1000 ,10 ,10 )
419+ ```
414420
415- data_model.add_meta_data(' X_shape' , (1000 , 10 , 10 ))
421+ write something to the meta data
422+ ``` python
423+ data_model.add_meta_data(' dimension' , (1000 , 10 , 10 ))
416424data_model.meta_data
417-
425+ >> > { ' dimension ' : ( 1000 , 10 , 10 )}
418426```
419427
420428# Anomaly detection (partial finished)
@@ -463,11 +471,15 @@ predictor.predict(AnomalyDataModel(X_test))
463471
464472# Data processing pipeline
465473
466- easy configuration, register steps and tasks in whole pipeline
474+ it features in the following:
475+ - easy configuration
476+ - register steps, tasks in data processing pipeline
477+ - log data result in each tasks, each steps
478+ - record the flow of pipeline, from steps to work (from marco to micro)
479+
480+ you can embed other function features in the task, but parameter: "data" is required to be passed in
467481
468- only for data processing
469- only for data processing
470- only for data processing
482+ e.g. add email notification, add log, upload to database etc...
471483
472484``` python
473485
0 commit comments