Working with multiple scenarios
You can generate multiple scenarios to test your model against a wide range of data and understand how robust the model is.
This example steps you through the process to generate multiple scenarios with a model. This makes it possible to test the performance of the model against multiple randomly generated data sets. It's important in practice to check the robustness of a model against a wide range of data. This helps ensure that the model performs well in potentially stochastic real-world conditions.
The example is the
StaffPlanning model in the
The example is structured as follows:
- The model
StaffPlanningcontains a default scenario based on two default data sets, along with five additional scenarios based on randomized data sets.
- The Python Notebook
CopyAndSolveScenarioscontains the random generator to create the new scenarios in the
For more information on working with scenarios, see the dd-scenario Python API documentation.