adpeps.ipeps.config¶
Configuration module for iPEPS simulations
These settings will be loaded from a configuration file .yml
file
via the from_dict()
function
-
D
: int = None¶ iPEPS bond dimension
-
base_sim
: Optional[str] = None¶ Base simulation
-
chi
: int = None¶ CTM boundary bond dimension
-
ctm_conv_tol
: float = 1e-10¶ CTM convergence criterium (singular values norm difference)
-
ctm_max_iter
: int = 20¶ Maximal number of CTM steps
-
ctm_min_iter
: int = 5¶ Minimal number of CTM steps
-
disp_level
: int = 1¶ no output)
- Type
Display level (0
-
flush_output
: bool = False¶ Passes the
flush = True
argument to the builtinprint
function when calling theadpeps.utils.printing.print()
functionUseful when deploying the code to computing clusters and capturing the output into text files
-
load_sim
: bool = None¶ Load previous (Python) simulation
-
max_iterations
: int = 100¶ Maximum number of optimizer iterations
-
method
: str = None¶ Optimization method
-
model
: str = None¶ Model
-
model_params
: dict = None¶ Model parameters
-
momentum_path
: str = 'Bril1'¶ Momentum path through the BZ
-
out_prefix
: str = ''¶ Optional prefix for the output file of the simulation
Example
.../{model}_D{D}_X{chi}.npz
becomes
.../{out_prefix}_{model}_D{D}_X{chi}.npz
-
pattern
: Optional[Iterable] = None¶ Unit cell configuration Defined as a 2-D array of integers that label the unique sites in the unit cell.
Example
A 2x2 unit cell with a [AB, BA]-type pattern is defined by
pattern = [ [0, 1], [1, 0] ]
-
resume
: bool = False¶ Resume earlier simulation if found
-
seed
: int = 1¶ Random seed for initial state