feynman-kac-python-prototype

A little prototype for a Feynman-Kac based Poisson solver for my bachelor thesis.

Installation

Assuming you have python and pip run first clone then in the root dir.

python -m venv venv
source venv/bin/activate
pip install -r requirements.txt

Useage

Run the python main.py in the src folder and any number of the optional arguments:

• -e, --epsilon defaults to .01 and sets the standard deviation that you want to reach
• -d, --dt0 defaults to .01 and sets the timestep at level 0 (or just the timestep for non-mlmc simulations)
• -x, --x defaults to .5 the starting x position
• -y, --y defaults to .5 the starting y position
• --non_homogeneous defaults to false uses the non-homogeneous test function
• -s, --standard_mc defaults to false if set runs non-mlmc simulation
• -N, --N_samples defaults to 256000 the number of samples for non-mlmc runs
• -w, --plot_walk defaults to false if set plots two correlated random walks
• -d, --debug defaults to false if set activates some more print statements

Jupyter notebooks

enter the virtual environment and start your jupyter lab or notebook from there