sdepack¶
Runge-Kutta numerical integration of scalar stochastic differential equations (SDEs) for Python.
Features¶
- Stochastic Runge-Kutta solvers (order 1 through 4) for time-invariant and time-variant scalar SDEs.
- Deterministic seeded random number generation via
R8_UNIFORMandR8_NORMAL(internal to solvers).
Mathematical model¶
Time-invariant routines integrate
\[
dX(t) = F(X)\,dt + Q\,G(X)\,dW(t),
\]
and time-variant routines integrate
\[
dX(t) = F(X,t)\,dt + Q\,G(X,t)\,dW(t),
\]
with uniform step size
\[
H = \frac{T_N - T_0}{N}.
\]
Stage noise variates follow
\[
W_i = Z_i\sqrt{\frac{Q_i Q}{H}},\quad Z_i \sim \mathcal{N}(0,1),
\]
with method-specific \(Q_i\) coefficients.
Public API¶
rk1_ti_solverk1_tv_solverk2_ti_solverk2_tv_solverk3_ti_solverk4_ti_solverk4_tv_solve
R8_UNIFORM and R8_NORMAL are internal Fortran helpers, not exported to Python.
Documentation¶
- Theory — numerical background and coefficients
- Quickstart — runnable examples
- API Reference — Fortran routine signatures
- References — source literature