kosmydel / public-transportation-solver

Genetic algorithm solving public transportation problem

Genetic algorithms

Description

Genetic algorithm solving public transportation problem. The goal is to find the best lines for public transportation.

---

Setup

1. Make sure you have Python 3.11 installed.
2. Install requirements

pip install -r requirements.txt

---

Running the app

The app consist of several modules:

• main.py - main module for running the app,
• grid_search.py - grid search for looking for hyperparameters,
• experiments.ipynb - notebook for testing random city generation and Cracow city graph.

Jupyter notebook

The easiest method to get some results is to run a Jupyter notebook server in the main repo's dir and open the experiments.ipynb file.

Main module

Description

This module shows a short example of how to use our genetic algorithm. It provides us with run_simulation function, which takes several parameters:

• G: Graph - graph representing the city,
• all_stops: list[int] - all stops in the city (vertices of the graph),
• best_paths - shortest paths between all stops
• no_of_generations: int - number of generations to simulate,
• report_every_n: int - report every n-th generation,
• report_show: bool - if True then the map image is displayed, otherwise it is saved to the file
• simulation_params: SimulationParams - optional simulation params

Running the app

To run the app, run the following command in repo's root directory:

PYTHONPATH="${PYTHONPATH}:$(pwd)/src" python src/main.py

The script will ask you to choose one of avaliable cities for simulations:

available simulations:
	[0] Random city
	[1] Cracow city
your choice:

After giving input to the program the calculations will begin.

Results

The graphs will be saved to results folder in the SVG format.

Grid search

Running the grid search

To run grid search, run the following command in repo's root directory:

PYTHONPATH="${PYTHONPATH}:$(pwd)/src" python src/grid_search.py

Setting the parameters

To set the parameters, edit the grid_search.py file. Look for variable named grid_search_params and edit it. Remember, that the grid search will run all possible combinations of the parameters. It may take a while.

Setting number of parallel processes

To set the number of parallel processes, edit the grid_search.py file. Look for variable named parallel_units and edit it. By default, it is set to -1, which means that the number of parallel processes will be equal to the number of cores on your machine.

Results

The results of the grid search will be saved in the results/gridsearch.csv file.

---

Development setup

Install hooks

pre-commit install

Running hook against all files

pre-commit run --all-files

---

Authors

• Jakub Kosmydel
• Norbert Morawski
• Bartłomiej Wiśniewski
• Przemysław Węglik