NonHomogeneousPoissonProcess

NonHomogeneousPoissonProcess represents a Poisson process with time-dependent intensity \(\lambda(t)\) and mean function

\[m(t)=\int_0^t \lambda(x)\,dx.\]

Constructor

NonHomogeneousPoissonProcess(intensity_function, mean_function=None)

Parameters

Parameter Description
intensity_function Callable intensity \(\lambda(t)\).
mean_function Optional callable mean function \(m(t)\).

Properties

Property Meaning
intensity_function The intensity callable.
mean_function The mean-function callable.

Methods

  • mean(t) — evaluate \(m(t)\).
  • count_probability(n, t) — compute the count probability at time t.
  • increment_probability(n, s, t) — compute the increment probability on [s, t].
  • simulate(t_max, rng=None) — generate events up to a finite horizon.

Example

from stochx.stochastic import NonHomogeneousPoissonProcess

process = NonHomogeneousPoissonProcess(
    intensity=lambda t: 1.0 + t,
    mean_function=lambda t: t + 0.5 * t**2,
)

print(process.intensity_function(2.0))
print(process.mean(2.0))
print(process.count_probability(2, 2.0))

PoissonProcess provides the homogeneous constant-rate case.