OnlineARIMA

class hana_ml.algorithms.pal.tsa.online_algorithms.OnlineARIMA(order=None, learning_rate=None, epsilon=None, output_fitted=True, random_state=None, random_initialization=None)

Online ARIMA implements an online learning method to estimate the parameters of ARIMA models by reformulating it into a full information online optimization task (without random noise terms), which has no limitations of depending on noise terms and accessing the entire large-scale dataset in advance.

Parameters:

order(p, d, m), tuple of int, optional

• p: value of the auto regression order.

• d: value of the differentiation order.

• m: extended order needed to transform all moving-averaging term to AR term.

Defaults to (1, 0, 0).

learning_ratefloat, optional

Learning rate. Must be greater than zero.

Calculated according to order(p, d, m).

epsilonfloat, optional

Convergence criterion.

Calculated according to learning_rate, p and m in the order.

output_fittedbool, optional

Output fitted result and residuals if True.

Defaults to True.

random_stateint, optional

Specifies the seed for random number generator.

• 0: use the current time (in second) as seed.

• Others: use the specified value as seed.

Default to 0.

random_initializationbool, optional

Whether randomly generate initial state

• False: set all to zero.

• True: use random values.

Defaults to False.

Examples

>>> arima = OnlineARIMA(order=(4,0,8), output_fitted=True, learning_rate=0.00001)
>>> res = arima1.partial_fit(data=df, key='TIMESTAMP')

Attributes:

model_DataFrame

Model content.

fitted_DateFrame

Fitted values and residuals.

Methods

partial_fit(data[, key, endog])	Generates a ARIMA model with given data.
predict([forecast_length, allow_new_index])	Generates time series forecasts based on the fitted model.
set_conn(connection_context)	Set connection context for an OnlineARIMA instance.

set_conn(connection_context)

Set connection context for an OnlineARIMA instance.

Parameters:

connection_contextConnectionContext

The connection to the SAP HANA system.

Returns:

None.

partial_fit(data, key=None, endog=None, **kwargs)

Generates a ARIMA model with given data.

Parameters:

dataDataFrame

DataFrame containing the data.

keystr, optional

The timestamp column of data. The type of key column is int.

Defaults to the first column of data if the index column of data is not provided. Otherwise, defaults to the index column of data.

endogstr, optional

The endogenous variable, i.e. time series.

Defaults to the first non-key column of data if not provided.

**kwargskeyword arguments, optional

The value of learning_rate and epsilon could be reset in the model.

For example, assume we have a OnlineARIMA object oa and we want to reset the value of learning_rate in the new training, then we can run

>>> oa.partial_fit(new_data, learning_rate=0.02)

Returns:

A fitted object of class "OnlineARIMA".

predict(forecast_length=None, allow_new_index=False)

Generates time series forecasts based on the fitted model.

Parameters:

forecast_lengthint, optional

Forecast horizon, i.e. number of future points to forecast.

Defaults to 1.

allow_new_indexbool, optional

Indicates whether a new index column is allowed in the result.

• True: return the result with new integer or timestamp index column.

• False: return the result with index column starting from 0.

Defaults to False.

Returns:

DataFrame

Prediction result, i.e. forecasted values within specified horizon, structured as follows:

• 1st column : timestamp

• 2nd column : forecast value

Inherited Methods from PALBase

Besides those methods mentioned above, the OnlineARIMA class also inherits methods from PALBase class, please refer to PAL Base for more details.