We will implement sampling from the parameter distributions obtained in Tasks 2a and 2b when running virtual experiments, and display the resulting distributions of protocol outputs, rather than single traces as at present. Some work on computational optimisation (including using faster simplifications of models known as emulators, and running on cloud and GPU resources) will be included to speed up this process, since many more simulations will need to be run than currently.
We will implement sampling from the parameter distributions obtained in Tasks 2a and 2b when running virtual experiments, and display the resulting distributions of protocol outputs, rather than single traces as at present. Some work on computational optimisation (including using faster simplifications of models known as emulators, and running on cloud and GPU resources) will be included to speed up this process, since many more simulations will need to be run than currently.