simpy

Use SimPy to define processes as Python generator functions that yield events like env.timeout(). Create an Environment, add shared resources such as simpy.Resource, start processes with env.process(), and run with env.run(until=time).

Create a simpy.Resource with the server capacity, then have each customer process request the resource using a with resource.request() context manager. A generator process yields exponential inter-arrival timeouts and spawns customer processes continuously.

Resource models limited-capacity servers, Container holds homogeneous bulk amounts like fuel with put/get by quantity, and Store queues discrete Python objects. FilterStore and PriorityStore add selective or priority-ordered retrieval.

Yes, simpy.rt.RealtimeEnvironment synchronizes simulation time with wall-clock time using a factor parameter, where factor=1.0 maps one simulation unit to one second. Strict mode raises errors when computation exceeds the real-time budget.

SimPy is not suited for continuous simulations with fixed time steps, which fit SciPy ODE solvers better. It also does not apply to independent processes without resource sharing or pure mathematical optimization problems.

Patch or subclass the resource's request and release methods to log queue lengths and timestamps, then compute time-weighted averages. The included ResourceMonitor class automates this and exports results to CSV.