simopt.curve_utils

Curve utility functions.

This module provides utility functions for manipulating and analyzing curves.

Module Contents

simopt.curve_utils.mean_of_curves(curves: collections.abc.Iterable[simopt.curve.Curve]) → simopt.curve.Curve

Compute the pointwise mean of a collection of curves.

All curves must have identical starting and ending x-values.

Parameters:

curves (Iterable[Curve]) – A collection of curves to aggregate.

Returns:

A curve representing the pointwise mean across all input curves.

Return type:

Curve

Raises:

TypeError – If the input is not an iterable of Curve objects.

simopt.curve_utils.quantile_of_curves(curves: collections.abc.Iterable[simopt.curve.Curve], beta: float) → simopt.curve.Curve

Compute the pointwise quantile of a collection of curves.

All curves must have identical starting and ending x-values.

Parameters:

  • curves (Iterable[Curve]) – A collection of curves to aggregate.

  • beta (float) – The quantile level to compute (e.g., 0.5 for median).

Returns:

A curve representing the pointwise quantile across the input curves.

Return type:

Curve

Raises:

TypeError – If input is not a valid collection of Curve objects.

simopt.curve_utils.cdf_of_curves_crossing_times(curves: collections.abc.Iterable[simopt.curve.Curve], threshold: float) → simopt.curve.Curve

Compute the CDF of crossing times from a collection of curves.

The crossing time is defined as the first x-value where a curve crosses a given threshold.

Parameters:

  • curves (list[Curve]) – A list of curves to analyze.

  • threshold (float) – The y-value at which to detect the first crossing.

Returns:

A curve representing the cumulative distribution of crossing times.

Return type:

Curve

Raises:

TypeError – If input is not a list of Curve objects.

simopt.curve_utils.quantile_cross_jump(curves: collections.abc.Iterable[simopt.curve.Curve], threshold: float, beta: float) → simopt.curve.Curve

Compute a curve with a jump at the quantile of the crossing times.

The curve is piecewise-constant with a single jump located at the specified quantile of the first crossing times across the input curves.

Parameters:

  • curves (list[Curve]) – A list of curves to analyze.

  • threshold (float) – The y-value at which to detect the first crossing.

  • beta (float) – The quantile level (e.g., 0.5 for median crossing time).

Returns:

A piecewise-constant curve with a jump at the quantile crossing time, if finite.

Return type:

Curve

Raises:

TypeError – If input types are incorrect.

simopt.curve_utils.difference_of_curves(curve_1: simopt.curve.Curve, curve_2: simopt.curve.Curve) → simopt.curve.Curve

Compute the difference between two curves (curve_1 - curve_2).

The x-values of both curves must align exactly.

Parameters:

  • curve_1 (Curve) – The first curve (minuend).

  • curve_2 (Curve) – The second curve (subtrahend).

Returns:

A curve representing the pointwise difference.

Return type:

Curve

Raises:

TypeError – If inputs are not Curve instances or are incompatible.

simopt.curve_utils.max_difference_of_curves(curve_1: simopt.curve.Curve, curve_2: simopt.curve.Curve) → float

Compute the maximum pointwise difference between two curves (curve_1 - curve_2).

Parameters:

  • curve_1 (Curve) – The first curve (minuend).

  • curve_2 (Curve) – The second curve (subtrahend).

Returns:

The maximum difference between the two curves at any x-value.

Return type:

float

Raises:

TypeError – If the inputs are not Curve instances or are incompatible.