Data generation for swarm simulation
[Modification is implementated in the original implementation of Siyu Zhou (https://github.com/siyuzhou)]
Two Boids with two obstacle and one goal data generation
python swarm_sim_data_generate.py B: No. of Boids #default:2
OBST: No. of obstacles #default:2
N_STEPS: No. of steps for which data is generated #default:200
Data can be generated for variable Boid and Obstacles to specific directory
python swarm_sim_data_generate.py --boids B --obstacles OBST --steps N_STEPS --save-dir /path/to/save/location/
*_edge.npy: Edge type data
*_time.npy: Time steps(dt)
*_timeseries.npy: N_STEPS x N x D tensor
N: Goal + No. of obstacles + No. of Boids ---- 1 + 2 + 2 D: State vector in form of (position, velocity) ---- 4 values
Edge types of the directed graph represent the influences between elements of the system.
Goal = 1, Obstacle = 2, Boid = 2
| |Goal|Obst|Obst 1|Obst 2|Boid 1|Boid 2|
|Goal | 0 | 0 | 0 | 0 | 1 | 1 |
|Obst 1| 0 | 0 | 0 | 0 | 2 | 2 |
|Obst 2| 0 | 0 | 0 | 0 | 2 | 2 |
|Boid 1| 0 | 0 | 0 | 0 | 0 | 3 |
|Boid 2| 0 | 0 | 0 | 0 | 3 | 0 |
