Python/maths/numerical_analysis/runge_kutta.py
Tianyi Zheng a8b6bda993
Delete arithmetic_analysis/ directory and relocate its contents (#10824)
* Remove eval from arithmetic_analysis/newton_raphson.py

* Relocate contents of arithmetic_analysis/

Delete the arithmetic_analysis/ directory and relocate its files because
the purpose of the directory was always ill-defined. "Arithmetic
analysis" isn't a field of math, and the directory's files contained
algorithms for linear algebra, numerical analysis, and physics.

Relocated the directory's linear algebra algorithms to linear_algebra/,
its numerical analysis algorithms to a new subdirectory called
maths/numerical_analysis/, and its single physics algorithm to physics/.

* updating DIRECTORY.md

---------

Co-authored-by: github-actions <${GITHUB_ACTOR}@users.noreply.github.com>
2023-10-23 09:31:30 +02:00

45 lines
1014 B
Python

import numpy as np
def runge_kutta(f, y0, x0, h, x_end):
"""
Calculate the numeric solution at each step to the ODE f(x, y) using RK4
https://en.wikipedia.org/wiki/Runge-Kutta_methods
Arguments:
f -- The ode as a function of x and y
y0 -- the initial value for y
x0 -- the initial value for x
h -- the stepsize
x_end -- the end value for x
>>> # the exact solution is math.exp(x)
>>> def f(x, y):
... return y
>>> y0 = 1
>>> y = runge_kutta(f, y0, 0.0, 0.01, 5)
>>> y[-1]
148.41315904125113
"""
n = int(np.ceil((x_end - x0) / h))
y = np.zeros((n + 1,))
y[0] = y0
x = x0
for k in range(n):
k1 = f(x, y[k])
k2 = f(x + 0.5 * h, y[k] + 0.5 * h * k1)
k3 = f(x + 0.5 * h, y[k] + 0.5 * h * k2)
k4 = f(x + h, y[k] + h * k3)
y[k + 1] = y[k] + (1 / 6) * h * (k1 + 2 * k2 + 2 * k3 + k4)
x += h
return y
if __name__ == "__main__":
import doctest
doctest.testmod()