NNEVE is a collection of neural network based solutions to physics based problems. As for now only network for quantum oscillator approximation is fully implemented. Hopefully soon will arrive neural network for solving Navier-Stokes equation based on limited number of measurement points..
This project is uploaded to PyPI as nneve, therefore can be installed with
following command
pip install nneveAt least Python 3.7 is required.
To view quantum oscillator approximation for states 1 to 7 you can load precalculated weights and acquire model object with following snippet:
from matplotlib import pyplot as plt
from nneve.quantum_oscillator.examples import default_qo_network
# acquire network object with precalculated weights
# for quantum oscillator state 1 (base)
network = default_qo_network(state=1)
network.plot_solution()
plt.plot()To manually run learning cycle check out "How to run QONetwork learning cycle" in Quantum Oscillator section of docs.
Online documentation is available at argmaster.github.io/NNEVE/
To build docs locally run
tox -e docs