splice

survival.splice(early, late, cutpoint)

Extrapolation: follow one curve to a cutpoint, then another’s shape.

Up to cutpoint, the result is early’s survival exactly. Beyond it, the result rescales late’s survival function to stay continuous at the cutpoint, early.survival(cutpoint) * late.survival(t) / late.survival(cutpoint), so the tail follows late’s hazard shape from that point on. This is the standard construction for a long-term model built on an observed curve that only covers a follow-up window: early is fit to the trial data, late is a parametric or external curve chosen for the extrapolation.

Parameters

Name	Type	Description	Default
early	SurvivalCurve	Curve followed up to cutpoint.	required
late	SurvivalCurve	Curve whose shape drives the extrapolation beyond cutpoint.	required
cutpoint	float	Time (years) at which the curve switches. Must be positive.	required

Returns

Name	Type	Description
	SurvivalCurve	A new SurvivalCurve that is continuous at cutpoint.

Example

from heormodel.survival import exponential, splice curve = splice(exponential(rate=0.1), exponential(rate=0.3), cutpoint=5.0) round(float(curve.survival(5.0)), 5) 0.60653