This notebook illustrates an end-to-end FP&A workflow in DataRobot. Time series forecasting in DataRobot has a huge suite of tools and approaches to handle highly complex multiseries problems.
DataRobot will be used for the model training, selection, deployment, and creating forecasts. While this example will leverage a snapshot file as a datasource this workflow applies to any data source, e.g. Redshift, S3, Big Query, Synapse, etc.
This notebook will demonstrate how to use the Python API client to:
- Connect to DataRobot
- Import and preparation of data for time series modeling
- Create a time series forecasting project and run Autopilot
- Retrieve and evaluate model performance and insights
- Making forward looking forecasts
- Evaluating forecasts vs. historical trends
- Deploy a model