FuzzyEntropy

class eyefeatures.features.measures.FuzzyEntropy(x=None, y=None, m=2, r=0.2, aoi=None, pk=None, return_df=True, ignore_errors=False)[source]

Bases: MeasureTransformer

Fuzzy Entropy.

A variation of Sample Entropy that uses a fuzzy membership function (typically exponential) to assess the similarity between vectors, rather than a hard Heaviside step function. This makes it more robust to noise and less sensitive to the choice of parameters. Measures the complexity of the scanpath.

Parameters:

  • m (int) – embedding dimension (length of sequences to compare).

  • r (float) – tolerance threshold/width of the fuzzy membership function (usually 0.2 * std).

  • x (str) – X coordinate column name.

  • y (str) – Y coordinate column name.

  • aoi (str) – Area Of Interest column name(-s).

  • pk (list[str]) – primary key.

  • return_df (bool) – whether to return output as DataFrame or numpy array.

  • ignore_errors (bool) – If True, return NaN when feature computation fails; otherwise raise.