auswahl.benchmarking.plot_selection

auswahl.benchmarking.plot_selection(pod: DataHandler, n_features: Union[int, Tuple[int, int]], dataset: Optional[str] = None, methods: Optional[Union[str, List[str]]] = None, plot_type: Literal['heatmap', 'bar'] = 'bar', save_path: Optional[str] = None)

Plots the selection probability for features of different selectors.

Parameters

pod: DataHandler

DataHandler object containing the benchmarking data.

n_features: Union[int, Tuple[int, int]]

feature configuration for which to plot the selection probability

dataset: str, default=None

Identifier of the dataset of which to plot the execution time. If there is data for only one dataset in the DataHandler object, the argument does not have to be specified.

methods: Union[str, List[str]], default=None

Method identifier or list of method identifiers for methods to be plotted. If None, all available methods are plotted.

plot_type: Literal[‘heatmap’, ‘bar’], default=’bar’

Plot type displayed. Option ‘heatmap’ is currently unavailable.

save_path: str, default=None

Path on which to store the plot. If None, the plot is simply displayed.