SARIMAX Learner (Labs)

Trains and generates a forecast using a Seasonal AutoRegressive Integrated Moving Average with eXogenous (SARIMAX) terms model. The SARIMAX models captures temporal structures in time series data in the following components:

  • AR (AutoRegressive): Relationship between the current observation and a number (p) of lagged observations.
  • I (Integrated): Degree (d) of differencing required to make the time series stationary.
  • MA (Moving Average): Time series mean and the relationship between the current forecast error and a number (q) of lagged forecast errors.

Seasonal versions of these components operate similarly, with lag intervals equal to the seasonal period (S).

Additionally, this node requires an exogenous column that externally influences the model. This column must be provided both for model training and forecasting. However, ensure that the number of rows in the exogenous variable to be used for forecasts must be equal to the number of forecasts to be made. Ensure that neither of the selected columns in the node configuration dialogue must contain a missing value.

Options

Sarimax Forecaster Settings

SARIMAX settings to configure the parameters for the model.

Target Column

Table containing training data for fitting the SARIMAX model, must contain a numeric target column with no missing values to be used for forecasting.

Exogenous Column

Table containing exogenous column for the SARIMAX model, must contain a numeric column with no missing values.

AR Order (p)

The number of lagged observations to be used in the model.

I Order (d)

The number of times to apply differencing before training the model.

MA Order (q)

The number of lagged forecast errors to be used in the model.

Seasonal AR Order (P)

The number of seasonally lagged observations to be used in the model.

Seasonal I Order (D)

The number of times to apply seasonal differencing before training the model.

Seasonal MA Order (Q)

The number of seasonal lagged forecast errors to be used in the model.

Seasonal Period (s)

Specify the length of the Seasonal Period

Log-transform data for modeling

Optionally log your target column before model fitting and exponentiate the forecast before output. This may help reduce variance in the training data.

Exogenous Column for Forecasting

Table containing exogenous column for making forecasts on the SARIMAX model, must contain a numeric column with no missing values and of length equal to the number of forecasts to be made.

Forecast

Forecasts of the given time series h period ahead of the training data.

Generate in-samples dynamically

Check this box to use in-sample prediction as lagged values. Otherwise use true values.

Input Ports

Icon

Table containing training data for fitting the SARIMAX model, must contain a numeric target column with no missing values to be used for forecasting. Additionally, this table must also contain the exogenous column to be used for training the SARIMAX model.

Icon

Table containing the exogenous column for forecasting on the SARIMAX model. It must contain a numeric column with no missing values.

Output Ports

Icon

Table containing forecasts for the configured column, the first value will be one timestamp ahead of the final training value used.

Icon

In sample model prediction values and residuals i.e. difference between observed value and the predicted output.

Icon

Table containing fitted model coefficients, variance of residuals (sigma2), and several model metrics along with their standard errors.

Icon

Pickled model object that can be used by the SARIMAX (Apply) node to generate different forecast lengths without refitting the model.

Popular Predecessors

  • No recommendations found

Popular Successors

  • No recommendations found

Views

This node has no views

Workflows

  • No workflows found

Links

Developers

You want to see the source code for this node? Click the following button and we’ll use our super-powers to find it for you.