Modeling and Simulation (CSCI 74000) Final Project

Authors: Artjom Plaunov and Daniel Mallia

Description

This repository constitutes an attempt to simulate outdoor (unowned / not pets) cat life in a fairly residential, urban environment for the purposes of understanding multiple trends and quantities with respect to the cats and their environmental impact. Namely, we simulate to understand:

• Cat growth trends, both the course of the simulation length and projected into the future by making use of the number of pregnancies at the end of the simulated period. In connection with this we estimate how many cats will be hit by cars.
• Mouse growth trends, understanding how the mouse population (explored here in simple fashion) may grow or shrink with respect to a certain cat population.
• The number of cat fights - a measure of how much of a nuisance cats might be to their human neighbors.

Simulation is conducted under an Agent Based Modeling (ABM) approach, making use of Python and the excellent ABM package, Mesa. We emphasize considerable temporal detail, making use of a 15 minute increment per step or tick in the simulation. We include use of 2 policies, so as to consider how humans might try to control the cat population:

• No control: in other words, no actions are taken against the cats.
• Removal: With a user-determined rate, cats are removed from the area.

Below is an example of the visual interface for our simulation:

Setup

This project assumes an installation of Python (>= 3.8.5) as well as the packages in requirements.txt. To install them you can run:

pip3 install -r requirements.txt

If any issues are encountered in installing using this requirments file (possibly due to cross-platform compatibility issues), this project really only requires 5 packages: mesa, pandas, numpy, scipy and matplotlib. As long as you can set up Python 3.8.5 (with which this was tested) and pip, then install these 5, you shouldn't encounter any problems.

How to run

To run the interactive version of the simulation use the following:

python3 server.py

Please note the following:

• server.py has multiple arguments for controlling which plots you would like to generate as the simulation runs. Pass --all_charts to include all, or choose among the other arguments.
• If you adjust parameters in the browser, such as the number of cats, you must reset the session (using the reset button on the top) for the changes to take effect.

To run the simulation in batch run mode for simulating growth trends and estimating parameters, use:

python3 batch_run.py

Please note the following:

• To have reproducible results, you should control the number of simulations to be run by using the --repro_iter and --seed arguments; for instance --repro_iter 10 --seed 1234 will execute 10 runs of the simulation under the given parameters and if run again for the same number of iterations with the same seed, the results will be identical. The default is to run 10 reproducible iterations with a seed of 1234.
• The default setting for --data_collection_period is 96; 96 ticks = 1 day under the initial simulation time setting of 15 minutes per tick. This will ensure that the growth trends will be updated on a per day basis. We advise leaving this as it is or increasing it for longer simulation runs; shortening it may cause a failure in the plotting of the growth trends