Before deploying the project, ensure the following libraries are installed using the specified commands:
- Keras:
pip install keras - Pandas:
pip install pandas - Scikit:
pip install scikit-learn - Tensor Flow:
pip install tensorflow - Virtual Environment:
pip install virtualenv - Pyglet version 1.5.11:
pip install pyglet==1.5.11 - Python Open GL:
pip install PyOpenGL
Download or clone the repository into a folder on your computer using the following link: https://github.com/MagmaArcade/Traffic-Flow-Prediction.git
To enable traffic prediction, train the learning models by running the following command for each training model, replacing {model_name} with options: "lstm," "gru," or "saes."
python train.py --model {model_name}
After training the models, run the program using the following command:
python gui.py
Arguments for gui.py --destination --origin --time --model
Example:
python gui.py --destination=3001 --origin=4201 --time=14:30 --model=gru
Alternatively, you can run the following command to get flow prediction for one node:
python task1-2test.py
Additional Arguments for task1-2test.py
--scats
--direction
--time
--model
Example:
python task1-2test.py --scats=3001 --direction=NE --time=22:15 --model=lstm
The program opens a GUI menu awaiting user input. Use the following keys to interact with the program:
Press the Space key to calculate routes. Press the Q key to cycle through calculated routes. Press Tab to toggle between selecting a new Origin or Destination. Left-click on a node to select a new Origin or Destination. Press the W key to cycle through models.
- xiaochus - Base code - Traffic Flow Prediction