sktime.transformations.panel.hog1d

class sktime.transformations.panel.hog1d.HOG1DTransformer(num_intervals=2, num_bins=8, scaling_factor=0.1)[source]

Bases: sktime.transformations.base._PanelToPanelTransformer

This class is to calculate the HOG1D transform of a dataframe of time series data. Works by splitting the time series num_intervals times, and calculate a histogram of gradients within each interval.

Parameters
  • num_intervals (int, length of interval.) –

  • num_bins (int, num bins in the histogram.) –

  • scaling_factor (float, a constant that is multiplied) – to modify the distribution.

transform(X, y=None)[source]

Function to transform a data frame of time series data.

Parameters

X (a pandas dataframe of shape = [n_samples, num_dims]) – The training input samples.

Returns

dims

Return type

a pandas data frame of shape = [n_samples, num_dims]