pricing.fit_references()

A model that reproduces every reference: expected layer losses (which

Usage

pricing.fit_references(
    layers=Ellipsis, frequencies=Ellipsis, default_alpha=2.0, rule="minimize"
)

may overlap or leave gaps) and excess frequencies.

Parameters

layers: list of tuple of float = Ellipsis

(limit, attachment, expected_loss) per layer.

frequencies: list of tuple of float = Ellipsis

(threshold, frequency) per excess frequency.

default_alpha: float = 2.0

Alpha above the highest point, unless an unlimited layer sets it.

rule: (minimize, midpoint) = "minimize"

Returns

TowerModel

Examples

>>> from prospicio.pricing import fit_references
>>> m = fit_references([(1000.0, 1000.0, 150.0), (3000.0, 1500.0, 160.0)], [(1000.0, 0.3)])
>>> round(m.layer_loss(3000.0, 1500.0), 6)

160.0