Statistical Forecast

The N2 Forecast module provides N2 nodes for creating statistical forecasts in the N2 environment for Odoo by exposing StatsForecast models. It enables data-driven forecasting based on historical time-series data and can be integrated into N2 workflows.

Dependencies

This module depends on the following modules *:

  • n2
  • n2_ui
  • n2_trigger
  • n2_data

*Built-in modules not listed

Core Forecast Nodes

ForecastDataNode

ForecastDataNode is responsible for extracting time-series data from Odoo models and transforming it into a DataFrame format compatible with StatsForecast models.

Parameters

  • key Raw
  • model Raw
  • order_by Raw
  • selected_fields Raw Eval

TrainNode

TrainNode is used to train a forecasting model using prepared historical data and a selected forecast model node.

Parameters

  • model_name Raw
  • freq Raw

ForecastNode

ForecastNode generates future predictions using a previously trained forecast model.

Parameters

  • forecast_model_id Raw
  • horizon Raw

ForecastModelStoreNode

ForecastModelStoreNode is used to persist trained models produced by the TrainNode.

ForecastResultStoreNode

ForecastResultStoreNode persists the forecast outputs generated by the ForecastNode.

Forecast Model Nodes

HistoricAverageNode

The HistoricAverageNode represents the Historic Average Model.

ArimaNode

The ArimaNode represent the ARIMA Model.

Parameters

  • p Raw
  • d Raw
  • q Raw
  • P Raw
  • D Raw
  • Q Raw
  • m Raw

AutoArimaNode

AutoArimaNode represents the AutoARIMA model, which automatically selects optimal ARIMA parameters.

Parameters

  • d Raw
  • D Raw
  • season_length Raw
  • approximation Raw
  • seasonal Raw

AutoEtsNode

AutoEtsNode represents the AutoETS (Error, Trend, Seasonality) model.

Parameters

  • season_length Raw
  • model Raw
  • damped Raw
  • phi Raw

AutoThetaNode

AutoThetaNode represents the AutoTheta forecasting model.

Parameters

  • season_length Raw
  • model Raw
  • decomposition_type Raw

Preprocessors

FillGapsNode

Use FillGapsNode fills missing timestamps in irregular time series.

Parameters

  • freq Raw

TimeAggregateNode

Use TimeAggregateNode aggregates time-series data into a coarser frequency.

Parameters

  • freq Raw
  • method Raw