make_landmarks_from_reference

ktch.datasets.make_landmarks_from_reference(reference: ndarray, n_samples: int = 30, sd: float = 1.0, random_state: int | RandomState | None = None, allow_collinearity: bool = False, allow_dup: bool = False, as_frame: bool = False) → ndarray | DataFrame

Generate landmark dataset around a reference configuration.

Parameters:

referencearray-like, shape (n_landmarks, n_dim)

Reference configuration.

n_samplesint, default=30

Number of samples to generate.

sdfloat, default=1.0

Standard deviation of the Gaussian noise added to the reference.

random_stateint, RandomState instance or None, default=None

Pass an int for reproducible output across multiple function calls.

allow_dupbool, default=False

If True, allow duplicate configurations in the generated dataset.

as_framebool, default=False

If True, return a pandas DataFrame.

Returns:

Xarray-like, shape (n_samples, n_landmarks, n_dim)

Generated landmark dataset.