stratify

stratify(lower, upper, bins, logarithmic=False, probability=False)

Stratify an axis into equal-width bins.

Mirrors the C# Stratify.XValues(StratificationOptions, isLogarithmic): splits [lower, upper] into bins equal-width strata (equal-width in log10 space when logarithmic=True), each carrying a weight defaulting to its own width. probability=True always returns zero rows, matching Stratify.XValues‘s own early return for probability-space options – the ported header exposes only the StratificationOptions overload used by the BestFit estimators’ profile-likelihood grids, not the probability-stratification methods (Probabilities, XToProbability, …), which are out of scope (see stratify.hpp’s file header).

lower, upper, and bins are validated rather than passed through to the C++ layer, which would otherwise silently return zero rows for lower >= upper (and for bins < 2) instead of raising an error.

Parameters

Name	Type	Description	Default
lower	float	Bounds of the axis to stratify. lower must be less than upper.	required
upper	float	Bounds of the axis to stratify. lower must be less than upper.	required
bins	int	Number of bins, greater than 1.	required
logarithmic	bool	Stratify on a log10 scale. Default False.	False
probability	bool	Mark the axis as a probability axis; kept only for parity with the C# constructor argument – it always yields zero bins (see above). Default False.	False

Returns

Name	Type	Description
	dict	Keys lower, upper, midpoint, weight (each :class:numpy.ndarray, one entry per bin).

Examples

>>> from corehydropy import stratify
>>> s = stratify(0, 1, bins=4)
>>> s["midpoint"]
array([0.125, 0.375, 0.625, 0.875])