represented_interval_weights

openstef_beam.metrics.metrics_helpers.represented_interval_weights(quantiles: Sequence[Quantile]) → ndarray[tuple[Any, ...], dtype[floating]]

Calculate the probability interval each quantile represents on [0, 1].

Interval boundaries are placed at the midpoints between consecutive quantiles, with the outer edges fixed at 0 and 1. Each quantile is weighted by the width of its interval.

Parameters:

quantiles (Sequence[Quantile]) – Quantile levels with shape (num_quantiles,). Must be sorted in ascending order and contain values in (0, 1).

Returns:

The interval weights with shape (num_quantiles,). Weights are non-negative and sum to 1.

Return type:

ndarray[tuple[Any, …], dtype[floating]]

Example

Unevenly spaced quantiles get different weights

>>> import numpy as np
>>> quantiles = np.array([0.05, 0.1, 0.5, 0.9, 0.95])
>>> represented_interval_weights(quantiles)
array([0.075, 0.225, 0.4  , 0.225, 0.075])

Parameters:

quantiles (Sequence[Quantile])

Return type:

ndarray[tuple[Any, ...], dtype[floating]]