Angular

polars_ts.imaging.angular

Gramian Angular Field (GAF) imaging for time series.

Converts time series to GASF (Gramian Angular Summation Field) and GADF (Gramian Angular Difference Field) images.

_min_max_scale(x)

Scale values to [-1, 1].

_paa(x, size)

Piecewise Aggregate Approximation — downsample by averaging segments.

_gasf_matrix(x, image_size)

Compute GASF for a single 1D series.

_gadf_matrix(x, image_size)

Compute GADF for a single 1D series.

to_gasf(df, image_size=None, id_col='unique_id', target_col='y')

Convert time series to Gramian Angular Summation Field images.

Rescales values to [-1, 1], converts to polar coordinates, and computes cos(phi_i + phi_j) for all pairs.

Parameters

df DataFrame with columns id_col and target_col. image_size Output image dimension. None for full resolution (n x n). Smaller values use Piecewise Aggregate Approximation (PAA). id_col Column identifying each time series. target_col Column with the time series values.

Returns

dict[str, np.ndarray] Mapping from series ID to a square 2D array with values in [-1, 1].

to_gadf(df, image_size=None, id_col='unique_id', target_col='y')

Convert time series to Gramian Angular Difference Field images.

Rescales values to [-1, 1], converts to polar coordinates, and computes sin(phi_i - phi_j) for all pairs.

Parameters

df DataFrame with columns id_col and target_col. image_size Output image dimension. None for full resolution (n x n). Smaller values use Piecewise Aggregate Approximation (PAA). id_col Column identifying each time series. target_col Column with the time series values.

Returns

dict[str, np.ndarray] Mapping from series ID to a square 2D array with values in [-1, 1].