yamle.utils.plotting_utils module

yamle.utils.plotting_utils.average_augmentations(result, metric)

This function will average the results of the augmentations and return the mean and standard deviation of the results.

It is performed across all levels of augmentation. For example, if we have 3 augmentations with 3 levels each, the function will return the mean and standard deviation of the 9 results.

Parameters:

• result (Dict[str, Any]) – The results dictionary

• metric (str) – The metric to average

Return type:

Union[Tuple[float, float], bool]

yamle.utils.plotting_utils.fetch_value_and_std_from_result(result, dataset, metric)

This is a helper function that will fetch the value and error from a result dictionary.

Return type:

Tuple[float, float]

yamle.utils.plotting_utils.pareto_optimal(values, metrics)

Selects the pareto optimal points from a list of points.

Parameters:

• values (List[List[Tuple[float, float]]]) – A list of value sets, where each value set contains a list of (mean, std dev) tuples.

• metrics (List[str]) – A list of metric names, one for each value set.

Return type:

Tuple[List[bool], List[namedtuple]]

Returns:

A tuple containing – - mask (List[bool]): A list indicating which points are pareto optimal - solutions (List[namedtuple]): The pareto optimal solutions