Figure out where to put adjust_inputs

[SeanMcOwen]

A lot of these can probably be state preparation functions to more easily encapsulate

def adjust_inputs(seed=0, upper_limit=1000): base_width = 4 base_height = 4 base_weight = 2 base_num_riders = 4

  seed = max(0, min(seed, 100))
  random.seed(time.time())

  increase_factor = 1 + (seed / 100) * 500
  random_factor = lambda: random.uniform(0.9, 1.1)

  grid_width = ceil(base_width * increase_factor * random_factor())
  grid_height = ceil(base_height * increase_factor * random_factor())
  grid_weight = ceil(base_weight * increase_factor * random_factor())
  num_riders = ceil(base_num_riders * increase_factor * random_factor())

  rider_driver_delta = random.randint(-100, -1)
  num_drivers = max(1, min(num_riders + rider_driver_delta, upper_limit))

  grid_width = max(2, min(grid_width, 15))
  grid_height = max(2, min(grid_height, 15))
  grid_weight = max(1, min(grid_weight, 10))
  num_riders = max(1, min(num_riders, upper_limit))
  num_drivers = max(1, min(num_drivers, num_riders - 1))

  grid = Grid(width=grid_width, height=grid_height)
  shortest_paths = grid.shortest_paths

  riders = [Rider(f'Rider {i}', random.choice(list(grid.zones.values()))) for i in range(num_riders)]
  drivers = [Driver(f'Driver {i}', random.choice(list(grid.zones.values())), {
      "cost_weight": 0.4, "distance_weight": 0.1, "duration_weight": 0.1,
      "driver_rating_weight": 0.2, "rider_rating_weight": 0.2, "acceptance_threshold": 20
  }) for i in range(floor(num_riders / 4))]

  ride_requests = {
      i + 1: RideRequest(i + 1, riders[i], random.choice(list(grid.zones.values())), random.choice(list(grid.zones.values())), grid)
      for i in range(num_riders)
  }

  return grid, riders, drivers, ride_requests