crm_perf_plots

This is script is used to obtain performance information of executing simulations given various configurations and plots the obtained results. It is meant to be run without any arguments and simply executes directly, storing information as image files.

>>> crm_perf_plots
run_sim(n_agents: int, domain_size: float, rng_seed: int = 0, n_threads: int = 1, n_voxels: int = 1) → tuple[float, int]
Parameters:

  • n_agents (int) – Number of initial agents

  • domain_size (float) – Domain size in x and y direction

  • rng_seed (int) – Seed to use for random generation of positions

  • n_threads (int) – Number of threads to use for simulation

  • n_voxesl (int) – Discretization of domain

Returns:

Wall time and number of final cells

Return type:

tuple[float, int]

../_images/computation-time-with-initial-agents.png

Runtime of the simulation depending on the number of initial agents. The domain size is kept constant thus increasing the density with the number of agents. The graph shows results with quadratic fits for 1,2,4 threads respectively.