Class Ctmc_transient_sens
Port of matlab/src/api/mc/ctmc_transient_sens.m, twin of
cpp/include/line/api/mc/ctmc_transient_sens.h and of the native
Python api.mc.ctmc_transient_sens. Differentiating the forward
equations d pi(t)/dt = pi(t) Q with respect to theta, with an initial vector
that does not depend on theta, gives Trivedi and Bobbio (2017), Eq. (9.82),
d/dt (dpi/dtheta) = (dpi/dtheta) Q + pi (dQ/dtheta), dpi(0)/dtheta = 0.
The sensitivity equation is DRIVEN by pi(t), so the two cannot be advanced
separately: state and sensitivity are integrated as ONE augmented system of
size 2n, which is also what keeps them consistent at every returned time
point. The integrator is the same LSODA that Ctmc_transient uses, so
the accepted grid is the JAR's own and need not coincide with the reference's
ode23 grid; the trajectories agree, the abscissae need not.
-
Nested Class Summary
Nested ClassesModifier and TypeClassDescriptionstatic final classDistribution and its sensitivity on the integrator's accepted grid. -
Method Summary
Modifier and TypeMethodDescriptionstatic Ctmc_transient_sens.Resultctmc_transient_sens(Matrix Q, Matrix dQ, double t1) From the uniform initial distribution over [0, t1].static Ctmc_transient_sens.Resultctmc_transient_sens(Matrix Q, Matrix dQ, Matrix pi0, double t1) From PI0 over [0, t1].static Ctmc_transient_sens.Resultctmc_transient_sens(Matrix Q, Matrix dQ, Matrix pi0, double t0, double t1)
-
Method Details
-
ctmc_transient_sens
From the uniform initial distribution over [0, t1]. -
ctmc_transient_sens
public static Ctmc_transient_sens.Result ctmc_transient_sens(Matrix Q, Matrix dQ, Matrix pi0, double t1) From PI0 over [0, t1]. -
ctmc_transient_sens
public static Ctmc_transient_sens.Result ctmc_transient_sens(Matrix Q, Matrix dQ, Matrix pi0, double t0, double t1) - Parameters:
Q- generator, n x ndQ- derivative of the generator with respect to theta, same sizepi0- initial distribution, 1 x nt0- initial timet1- final time
-