Statistical Features

The stats submodule provides statistical analysis of eye movement components.

Base Feature Transformer

class eyefeatures.features.stats.StatsTransformer(features_stats, x=None, y=None, t=None, duration=None, dispersion=None, aoi=None, calc_without_aoi=False, pk=None, return_df=True, warn=True)[source]

Bases: BaseTransformer

Base class for statistical features. Aggregate function strings must be compatible with pandas. Expected dataframe with fixations.

Parameters:

  • features_stats (dict[str, list[str]]) – Dictionary of format {‘feature_1’: [‘statistic_1’, ‘statistic_2’], …}.

  • x (str) – X coordinate column name.

  • y (str) – Y coordinate column name.

  • t (str) – timestamp column name.

  • duration (str) – duration column name (milliseconds expected).

  • dispersion (str) – fixation dispersion column name.

  • aoi (str | list[str]) – Area Of Interest column name(-s). If provided, features can be calculated inside the specified AOI.

  • calc_without_aoi (bool) – if True, then, in addition to AOI-wise features, calculate regular features ignoring AOI.

  • pk (list[str]) – primary key.

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

  • warn (bool) – whether to enable warnings.

get_feature_names_out(input_features=None)[source]

Returns the names of the features generated by this transformer.

Return type:

list[str]

Feature Transformers