GramianAngularField

class pyts.image.GramianAngularField(image_size: int | float = 1.0, sample_range: tuple[float, float] | None = (-1, 1), method: str = 'summation', overlapping: bool = False, flatten: bool = False)

Gramian Angular Field.

Parameters:

image_sizeint or float (default = 1.)

Shape of the output images. If float, it represents a percentage of the size of each time series and must be between 0 and 1. Output images are square, thus providing the size of one dimension is enough.

sample_rangeNone or tuple (min, max) (default = (-1, 1))

Desired range of transformed data. If None, no scaling is performed and all the values of the input data must be between -1 and 1. If tuple, each sample is scaled between min and max; min must be greater than or equal to -1 and max must be lower than or equal to 1.

method‘summation’ or ‘difference’ (default = ‘summation’)

Type of Gramian Angular Field. ‘s’ can be used for ‘summation’ and ‘d’ for ‘difference’.

overlappingbool (default = False)

If True, reduce the size of each time series using PAA with possible overlapping windows.

flattenbool (default = False)

If True, images are flattened to be one-dimensional.

References

[1]

Z. Wang and T. Oates, “Encoding Time Series as Images for Visual Inspection and Classification Using Tiled Convolutional Neural Networks.” AAAI Workshop (2015).

Examples

>>> from pyts.datasets import load_gunpoint
>>> from pyts.image import GramianAngularField
>>> X, _, _, _ = load_gunpoint(return_X_y=True)
>>> transformer = GramianAngularField()
>>> X_new = transformer.transform(X)
>>> X_new.shape
(50, 150, 150)

Methods

__init__([image_size, sample_range, method, ...])
fit([X, y])	Pass.
fit_transform(X[, y])	Fit to data, then transform it.
get_metadata_routing()	Get metadata routing of this object.
get_params([deep])	Get parameters for this estimator.
set_params(**params)	Set the parameters of this estimator.
transform(X)	Transform each time series into a GAF image.

__init__(image_size: int | float = 1.0, sample_range: tuple[float, float] | None = (-1, 1), method: str = 'summation', overlapping: bool = False, flatten: bool = False) → None

fit(X: ArrayLike | None = None, y: ArrayLike | None = None) → Self

Pass.

Parameters:

X

Ignored

y

Ignored

Returns:

selfobject

transform(X: ArrayLike) → NDArray[float64]

Transform each time series into a GAF image.

Parameters:

Xarray-like, shape = (n_samples, n_timestamps)

Returns:

X_newarray-like, shape = (n_samples, image_size, image_size)

Transformed data. If flatten=True, the shape is (n_samples, image_size * image_size).

Examples using pyts.image.GramianAngularField

Data set of Gramian angular fields

Data set of Gramian angular fields

Single Gramian angular field

Single Gramian angular field