Can't define unique characteristics for each locust

[luojxxx]

Description of issue / feature request

I would like to simulate a group of GPS tracked animals by using init within the BehaviorClass to instantiate a Animal class which will iterate through GPS coordinates upon every method call that will be POSTed to the API.

Expected behavior

I expected to be able to use def init(self): to instantiate an DeviceTracker class object with a get_position_update method that returns the GPS coordinates (method acts like a generator that's looped on itself).

Actual behavior

TypeError: init() takes 1 positional argument but 2 were given

Environment settings (for bug reports)

• OS: Mac OS 10.13.3
• Python version: 3.5.3
• Locust version: 0.8.1

Steps to reproduce (for bug reports)

class TrackerBehavior(TaskSet):
  def __init__(self):
    self.tracker = DeviceTracker(
      selected_device_id, 
      selected_animal_id, 
      selected_path, 
      selected_starting_point
      )

  @task(1)
  def update_position(self):
    self.client.post("/api/v1/position_tracking", self.tracker.get_position_update())

[luojxxx]

Figured it out. Needed to instantiate the DeviceTracker using the on_start method, and then the DeviceTracker class itself needs to select or randomly select the particular animal to simulate.

class TrackerBehavior(TaskSet):
  def on_start(self):
    self.locust.tracker = DeviceTracker()

  @task(1)
  def update_position(self):
    self.client.post("/api/v1/position_tracking", self.locust.tracker.get_position_update())