# Simulation parameters
MAX_STEPS = 100
MAX_PARTICLES = 300
TRIALS_PARTICLES = 10 # Number of trials per particle size experiment
OUTFILE = 'data/mcpomdp_metrics.csv'
GUI_ON = False
#Metrics - num_particles, iterations, time, goal_dist, error_mean, error_std
printer(['particles','iters','rtime','goaldist','errormean','errorstd'], OUTFILE)
for num_particles in np.arange(1,MAX_PARTICLES):
for _ in range(TRIALS_PARTICLES):
sim_params['num_particles'] = num_particles
# Initialize model of world
sim = Simulation(world_params)
sstate = sim.get_state()
# Reinitialize randomness after world generation
np.random.seed(None)
# Add a car model robot
init_state = [5,5,0]
sim.add_robot(robot_params,init_state)
# add a radar sensor
robot_id = robot_params['id']
sim.add_sensor(robot_id, sensor_params)
readings = sim.read_sensors()
from model.Simulation import Simulation
from view.render2d import View
#import matplotlib.pyplot as plt
#import numpy as np
sim_params = {'world_type':'maze',
'world_length':100,
'world_width':100,
'robot_length':2,
'robot_stochastic':False,
'radar_radius':8,}
robot_id = 1
v = View()
sim = Simulation(sim_params)
state = sim.get_state()
v.update_rstate(state)
sim.add_robot('car',robot_id)
state = sim.get_state()
v.update_rstate(state)
sim.add_sensor(robot_id, 'radar')
readings = sim.read_sensors()
print readings
raw_input('pause : press any key to finish...')
#maze = gen_terrain_dfs(100,100)
#world = Maze()
