Example: GrenobleSand (synthetic Van Genuchten → neural fit)
This example generates synthetic data from a VanGenuchten model (“Grenoble sand” parameters), then fits a neural PINN to recover the parameters and compares reference vs. prediction.
- Reference model:
VanGenuchten(Ks, alpha, m, theta_range=(0, θs)) - Fit model:
VanGenuchten(Ks=Param, m=Param, theta_range=(0, θs)) - Workflow: forward solve → neural fit → plot comparison
How to run
From the command line: ```bash python -m frontx.examples.grenoblesand