How madagra/basic-pinn works
Basic PINN is an educational Python package for approximating solutions to differential equations. Users run either the logistic ODE or vibrating-string wave-equation example; each defines its equation loss and sampling procedure, trains a neural network, then plots or animates the result. Shared code provides the network and derivative functions, functional-optimizer training, and visualizations. The examples’ imports and calls are visible in the excerpts; the README documents how users launch them. No storage or external service is part of the runtime architecture.
Main entry points: logistic_equation_1d.py Logistic solver · wave_equation_1d.py Wave solver · pinn.py Neural network · training.py PINN training
Components 7
Equation examples
- Logistic solver
basic_pinn/logistic_equation_1d.py - Wave solver
basic_pinn/wave_equation_1d.py
PINN core
- Neural network
basic_pinn/pinn.py - Derivative functions
basic_pinn/pinn.py - PINN training
basic_pinn/training.py
Visualization
- Solution plots
basic_pinn/plotting.py
Other components
- User
Connections 10
- User → Logistic solver: runs example
README.md - User → Wave solver: runs example
README.md - Logistic solver → Neural network: creates model
basic_pinn/logistic_equation_1d.py - Logistic solver → Derivative functions: builds derivatives
basic_pinn/logistic_equation_1d.py - Logistic solver → PINN training: trains model
basic_pinn/logistic_equation_1d.py - Logistic solver → Solution plots: plots solution
basic_pinn/logistic_equation_1d.py - Wave solver → Neural network: creates model
basic_pinn/wave_equation_1d.py - Wave solver → Derivative functions: builds derivatives
basic_pinn/wave_equation_1d.py - Wave solver → PINN training: trains model
basic_pinn/wave_equation_1d.py - Wave solver → Solution plots: animates solution
basic_pinn/wave_equation_1d.py