sdeint is a collection of numerical algorithms for integrating Ito and Stratonovich stochastic ordinary differential equations (SODEs). It has simple functions that can be used in a similar way to scipy.integrate.odeint()
or MATLAB's ode45
.
There already exist some python and MATLAB packages providing Euler-Maruyama and Milstein algorithms, and a couple of others. So why am I bothering to make another package?
It is because there has been 25 years of further research with better methods but for some reason I can't find any open source reference implementations. Not even for those methods published by Kloeden and Platen way back in 1992. So I will aim to gradually add some improved methods here.
This is prototype code in python, so not aiming for speed. Later can always rewrite these with loops in C when speed is needed.
Warning: this is an early pre-release. Wait for version 1.0. Bug reports are very welcome!
itoint(f, G, y0, tspan)
for Ito equation dy = f(y,t)dt + G(y,t)dWstratint(f, G, y0, tspan)
for Stratonovich equation dy = f(y,t)dt + G(y,t)∘dWThese work with scalar or vector equations. They will choose an algorithm for you. Or you can use a specific algorithm directly:
itoEuler(f, G, y0, tspan)
: the Euler-Maruyama algorithm for Ito equations.stratHeun(f, G, y0, tspan)
: the Stratonovich Heun algorithm for Stratonovich equations.itoSRI2(f, G, y0, tspan)
: the Rößler2010 order 1.0 strong Stochastic Runge-Kutta algorithm SRI2 for Ito equations.itoSRI2(f, [g1,...,gm], y0, tspan)
: as above, with G matrix given as a separate function for each column (gives speedup for large m or complicated G).stratSRS2(f, G, y0, tspan)
: the Rößler2010 order 1.0 strong Stochastic Runge-Kutta algorithm SRS2 for Stratonovich equations.stratSRS2(f, [g1,...,gm], y0, tspan)
: as above, with G matrix given as a separate function for each column (gives speedup for large m or complicated G).stratKP2iS(f, G, y0, tspan)
: the Kloeden and Platen two-step implicit order 1.0 strong algorithm for Stratonovich equations.deltaW(N, m, h)
: Generate increments of m independent Wiener processes for each of N time intervals of length h.Ikpw(dW, h, n=5)
: Approximate repeated Ito integrals.Jkpw(dW, h, n=5)
: Approximate repeated Stratonovich integrals.Iwik(dW, h, n=5)
: Approximate repeated Ito integrals.Jwik(dW, h, n=5)
: Approximate repeated Stratonovich integrals.import numpy as np import sdeint a = 1.0 b = 0.8 tspan = np.linspace(0.0, 5.0, 5001) x0 = 0.1 def f(x, t): return -(a + x*b**2)*(1 - x**2) def g(x, t): return b*(1 - x**2) result = sdeint.itoint(f, g, x0, tspan)
x = (x1, x2)
, dW = (dW1, dW2)
and with initial condition x0 = (3.0, 3.0)
import numpy as np import sdeint A = np.array([[-0.5, -2.0], [ 2.0, -1.0]]) B = np.diag([0.5, 0.5]) # diagonal, so independent driving Wiener processes tspan = np.linspace(0.0, 10.0, 10001) x0 = np.array([3.0, 3.0]) def f(x, t): return A.dot(x) def G(x, t): return B result = sdeint.itoint(f, G, x0, tspan)
itoEuler
:stratHeun
:itoSRI2, stratSRS2
:stratKP2iS
:Ikpw, Jkpw
:Iwik, Jwik
:- Rewrite
Iwik()
andJwik()
so they don't waste so much memory. - Fix
stratKP2iS()
. In the unit tests it is currently less accurate thanitoEuler()
and this is likely due to a bug. - Implement the Ito version of the Kloeden and Platen two-step implicit alogrithm.
- Add more strong stochastic Runge-Kutta algorithms. Perhaps starting with Burrage and Burrage (1996)
- Currently prioritizing those algorithms that work for very general d-dimensional systems with arbitrary noise coefficient matrix, and which are derivative free. Eventually will add special case algorithms that give a speed increase for systems with certain symmetries. That is, 1-dimensional systems, systems with scalar noise, diagonal noise or commutative noise, etc. The idea is that
itoint()
andstratint()
will detect these situations and dispatch to the most suitable algorithm. - Eventually implement the main loops in C for speed.
- Some time in the dim future, implement support for stochastic delay differential equations (SDDEs).
nsim
: Framework that uses this sdeint
library to enable massive parallel simulations of SDE systems (using multiple CPUs or a cluster) and provides some tools to analyze the resulting timeseries. https://github.com/mattja/nsim