data.origami_sampler.md

data.origami_sampler

Sampling utilities for generating origami fold tasks.

Classes

OrigamiSampler

Stateless sampler that produces a folded paper trajectory summary.

Parameters

Name	Type	Default	Description
max_fold_attempts	int	20	Maximum random fold-line attempts per fold step.
max_paper_retries	int	10	Maximum retries at each step before moving on.

Methods

sample(level: int) -> dict[str, Any]

Generates a folded paper task with up to level successful folds.

Parameters

Name	Type	Description
level	int	Maximum number of fold steps to attempt.

Returns

Returns a dictionary with the following keys:

Key	Type	Description
total_action	np.ndarray	Flattened fold endpoints [x1, y1, x2, y2, ...]; shape (4 * actual_folds,).
actual_folds	int	Number of successful folds.
final_paper	paper.Paper	Folded paper state after sampling.

Examples

from data.origami_sampler import OrigamiSampler

sampler = OrigamiSampler()
sample = sampler.sample(level=3)
print(sample["actual_folds"], sample["total_action"].shape)

Expected output (example):

3 (12,)

Notes

• actual_folds <= level.
• If no fold succeeds, total_action is an empty array with shape (0,).