How to use mesa_geo and have the agents compute path and move on it?

[PreetamKulkarni]

I have a crowd-logistics model (created using mesa) in which I used continuous space to create my agents . These agents are supposed to move to a destination as needed in each tick. To make my model more realistic, I need to put these agents on an actual map and have them use actual routes. Is this possible?

I saw in another discussion on this group (using only routes #162) that referred to Agents and Networks example. I tried to down load this model but I get the following error:
ModuleNotFoundError: No module named 'src'

I have maintained the same folder structure as shown in the example.

[wang-boyu]

Sorry about the confusion. As mentioned in the README file, in the Agents and Networks example folder, you'll need to do a

python3 -m pip install -r requirements.txt

so that the src module can be installed in the virtual environment (if you use a virtual environment). Then the model should be able to run or be imported.

Let me know if this works.

[PreetamKulkarni]

Hi @wang-boyu,

Thank you for the response. I followed the instructions. I used the command you showed above and installed the packages through the requirement.txt file.

However, I still get thf following error:

usage: Agents and Networks in Python [-h] --campus CAMPUS
Agents and Networks in Python: error: the following arguments are required: --campus
An exception has occurred, use %tb to see the full traceback.

SystemExit: 2

[wang-boyu]

Hi, the model should be run via the following command (in the example directory):

python3 scripts/run.py --campus ub

or change ub to gmu for a different campus map:

python3 scripts/run.py --campus gmu

Hopefully this will get the model running : )

[PreetamKulkarni]

Hi,
Thanks for sharing this info. I was trying to run the run.py directly and that's why it was not working.

It opens the interface now.

However, after the first step, it stops and gives the following error:

!python3 scripts/run.py --campus ub Interface starting at http://127.0.0.1:8521 Socket opened! {"type":"reset"} {"type":"reset"} {"type":"get_step","step":1} {"type":"get_step","step":2} Uncaught exception GET /ws (127.0.0.1) HTTPServerRequest(protocol='http', host='127.0.0.1:8521', method='GET', uri='/ws', version='HTTP/1.1', remote_ip='127.0.0.1') Traceback (most recent call last): File "/Users/preetamkulkarni/anaconda3/lib/python3.11/site-packages/tornado/websocket.py", line 630, in _run_callback result = callback(*args, **kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Users/preetamkulkarni/anaconda3/lib/python3.11/site-packages/mesa_viz_tornado/ModularVisualization.py", line 230, in on_message self.application.model.step() File "/Users/preetamkulkarni/Library/CloudStorage/OneDrive-UniversityofTexasatArlington/NSF CAREER - Project/Python MESA/Agents and Networks/src/model/model.py", line 212, in step self.schedule.step() File "/Users/preetamkulkarni/anaconda3/lib/python3.11/site-packages/mesa/time.py", line 145, in step self.do_each("step", shuffle=True) File "/Users/preetamkulkarni/anaconda3/lib/python3.11/site-packages/mesa/time.py", line 121, in do_each getattr(self._agents[agent_key], method)() File "/Users/preetamkulkarni/Library/CloudStorage/OneDrive-UniversityofTexasatArlington/NSF CAREER - Project/Python MESA/Agents and Networks/src/agent/commuter.py", line 88, in step self._prepare_to_move() File "/Users/preetamkulkarni/Library/CloudStorage/OneDrive-UniversityofTexasatArlington/NSF CAREER - Project/Python MESA/Agents and Networks/src/agent/commuter.py", line 134, in _prepare_to_move self._path_select() File "/Users/preetamkulkarni/Library/CloudStorage/OneDrive-UniversityofTexasatArlington/NSF CAREER - Project/Python MESA/Agents and Networks/src/agent/commuter.py", line 189, in _path_select self.model.walkway.cache_path( File "/Users/preetamkulkarni/Library/CloudStorage/OneDrive-UniversityofTexasatArlington/NSF CAREER - Project/Python MESA/Agents and Networks/src/space/road_network.py", line 87, in cache_path with open(self._path_cache_result, "wb") as cached_result: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ FileNotFoundError: [Errno 2] No such file or directory: 'outputs/ub_path_cache_result.pkl'

Iget a similar error with "!python3 scripts/run.py --campus gmu"

[wang-boyu]

Yeah sorry this is a bug that just got fixed. Do you mind pulling latest code and try again?

[wang-boyu]

Or else you could also create an empty folder named outputs in the example folder. This should fix it.

[PreetamKulkarni]

Hi @wang-boyu,

Thank you for fixing the issue. It is working on my computer now.

I have a crowd-logistics model (created using mesa) in which I used continuous space to create my agents . These agents are supposed to move to a destination as needed in each tick. To make my model more realistic, I need to put these agents on an actual map and have them use actual routes. Is this possible based on your experience with mesa-geo? I am trying to understand how hard this would be to achieve.

[wang-boyu]

Yes this is exactly what happens in the agents-and-networks example model. Commuters' __path_select() function computer their shortest path on a road network, which is implemented as a separate class here. In this model all agents move at the same speed, and there's no agent-agent interactions (e.g., speed doesn't change if there's traffic jam ahead). You could add some heterogeneity by letting agents having different speed, whereas agent-agent interaction would require more effort to update the model.

[PreetamKulkarni]

Thank you!