Model technical summary

The model is written in the python programming language and uses a simulation package called simpy. The model code is organised in a Jupyter notebook. The notebook is organised as follows:

Imports

The code imports the python libraries needed to run the computer simulation.

Constants and defaults for modelling as-is

This allows a user to set the default distribution parameters, time-dependent arrival rate, resource counts, and simulation settings (number of replications, run length etc.)

Utility functions

Non-core code that provides utilities for simulation model users.

Distribution classes

High level distribution classes built on numpy with their own random number stream.

Model Parameterisation

A scenario class that is initialised using user set defaults and can be modified to execute different experiments. The scenario class also controls common random numbers.

Patient Pathways Process Logic

This section contains the actual model logic. It is organised into two classes to mimic the logic for trauma and non-trauma patients.

Main model

This section creates the model and generates random patient arrivals to the model.

Logic to process end of run results

This section contains code to produce a summary of simulation results

Executing a model

Contains code demonstrating how to perform a base run (and multiple replications) of the model and obtain and display results.

Scenario analysis

Contains full set of experiments performed with the model. The code creates the different experiments and run each in turn. A summary of results is provided a the end.

Script to produce formatted LaTeX table for paper

Recreates the LaTeX used to report the results as a table in a paper.