New answers tagged runge-kutta
2
The first method is Euler-Maruyama with its strong convergence of order $0.5$. This can be seen in the difference plot on the right. The step sizes vary with a factor of $4$, while the error does not decrease that rapidly.
In the second method, if one inserts the constants actually used, then the steps are
\begin{align}
k_1&=f(x_t)Δt+g(x_t)\sqrt2ΔW_t\\
...
Top 50 recent answers are included
Related Tags
runge-kutta × 132ode × 55
numerics × 35
python × 18
time-integration × 14
computational-physics × 11
c++ × 10
finite-difference × 9
differential-equations × 9
pde × 8
stability × 7
eigenvalues × 6
reference-request × 6
integration × 6
matlab × 5
nonlinear-equations × 5
error-estimation × 5
fluid-dynamics × 4
scipy × 4
solver × 4
adaptive-timestepping × 4
linear-algebra × 3
algorithms × 3
simulation × 3
fortran × 3