adpeps.simulation.run_ipeps_gs¶

Main ground-state executable script

Note

The simulations are intended to be used by calling the package directly via python -m adpeps ..., as described in Getting Started

class Objective(peps)[source]¶

Class containing the objective function for the optimizer

Parameters: peps (adpeps.ipeps.ipeps.iPEPS) – iPEPS object to be optimized

jac(v)[source]¶

Computes the vector-Jacobian product (gradient) of the iPEPS at the current point v in parameter space

This function is designed to work with optimizers that make separate calls to the objective function and the gradient, by caching both.

Parameters: v (jax._src.numpy.lax_numpy.ndarray) – input variational parameters for the site tensors of the iPEPS
Returns: gradient of the iPEPS with respect to the input parameters v
Return type: ndarray

out(v)[source]¶

Computes the objective function (energy) of the iPEPS

Since many optimizers make separate calls to the objective function and the gradient, but usually require both results for any given parameter vector v, this function calls the gradient as well.

Parameters: v – input variational parameters for the site tensors of the iPEPS
Returns: energy of the iPEPS at the point v in parameter space

fun¶: Objective function - CTM iterations until convergence followed by the evaluation of the energy

property gradnrm¶: Norm of the gradient

peps¶: iPEPS object

run(config_file)[source]¶

Start the simulation

Parameters: config_file (str) – filename of the configuration file