Calibrating the Virtual Ecosystem

This section aims to explain what to do if you experience unexpected values in outputs or crashes from running the Virtual Ecosystem. The best approach to take depends on exactly what the issue you are running into is:

• if the model runs successfully, but the values it is producing make no sense then we would recommend first trying to identify the bug by plotting the simulation output.

• if the model crashes, but does so in a gracefully manner (i.e. generates easy to read log messages rather than a complex exception trace), then the simulation log is the first place to look, as it should give specific information about what went wrong. Often that may point towards an input that has unexpected values and where fixing the inputs will resolve the problem.

• if the steps above don’t resolve your issue or the model crash produces an exception trace rather than a log message, you will need to debug the model - watch it running and see what variables change when and how these might lead to errors.