odeint

package
v0.0.0-...-45afac1 Latest Latest
Warning

This package is not in the latest version of its module.

Go to latest
Published: Jun 16, 2018 License: MIT Imports: 1 Imported by: 4

Documentation

Overview

Package odeint implements Ordinary Differential Equations integrators. for initial value problems to be solved by explicit methods. The package features methods for the standard library types,

float32
float64
complex64
complex128

Methods for float64

The import to use float64 is

import "github.com/Daniel-M/odeint/float64"

The integrator methods implemented so far are,

* Euler

* Mid point

* Runge-Kutta 4

The package is easily extensible to provide other methods, you can follow the template files as reference,

templates/stepper_method.go.t
templates/stepper_method_test.go.t

Example - Simple harmonic oscillator

The integrator can be used to integrate the ODE for the harmonic oscillator.

x'' + p*x' + k*x = 0

which can be decomposed as the system,

x' = u
u' = -p*u - k*x

If we want to solve the system using float64, we must import the adequate subpackage,

import "github.com/Daniel-M/odeint/float64"

So we begin by defining the system of coupled differential equations

func odesys(x []float64, parameters []float64) []float64 {

  dxdt := make([]float64, len(x))

  dxdt[0] = x[1]
  dxdt[1] = -parameters[0]*x[0] - parameters[1]*x[1]

  return dxdt
}

declare the state and parameters variables,

state := make([]float64, 2)
params := make([]float64, 2)

Putting the inital conditions

state[0] = 0.2
state[1] = 0.8

And the parameters

params[0] = 1.2 * 1.2
params[1] = 0.2

We create an instance of the system,

system := odeint.NewSystem(state, params, odesys)

And an instance of the integrator with Midpoint method,

var integrator odeint.Midpoint

Set the system to the integrator before integrating the system

err := integrator.Set(0.1, *system)
if err != nil {
  panic(err)
}

And finally we integrate within a loop

for i := 0; i < int(30.0/integrator.StepSize()); i++ {
  fmt.Println(float64(i)*integrator.StepSize(), state)

  state, err = integrator.Step()
  if err != nil {
    panic(err)
  }
}

The code above will print the data columns to the standard output. To write to a file you could create a file with

os.Create

and write to it with

fmt.Fprintf(w,...)

where w implements the interface

io.Writter

There are more examples at the examples path

FAQ

A subpackage for each numeric type?

You might be thinking, why does this guy have a subpackage for each numeric type? Well, though it makes the package harder to maintain, having type specific integrators is a priority for me. I could have used interface-based integrators but it would be at the expense of the extensibility of the integrators to more custom numerical types, a feature which I find relevant too.

License

This code is licensed under MIT license that can be found in the LICENSE file as,

MIT License

Copyright (c) 2017-2018 Daniel Mejía Raigosa

Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.

And at the begining of the source code files as the fragment,

Copyright 2017-2018 Daniel Mejía Raigosa. All rights reserved.
Use of this source code is governed by a MIT
license that can be found in the LICENSE file.

Index

Constants

This section is empty.

Variables

This section is empty.

Functions

This section is empty.

Types

type Error

type Error struct {
	// contains filtered or unexported fields
}

Error implements an error custom type.

func (*Error) Error

func (e *Error) Error() string

Error implementation of the Error() interface for the package "error".

type Euler

type Euler struct {
	// contains filtered or unexported fields
}

Euler implements the Euler stepper method. Euler is part of the Stepper interface.

func NewEuler

func NewEuler(stepSize float64, system System) (r *Euler)

NewEuler returns a reference to a new Euler stepper method.

func (*Euler) Set

func (euler *Euler) Set(stepSize float64, system System) error

Set sets the step size and System for the Euler stepper.

func (*Euler) SetState

func (euler *Euler) SetState(state []float64) error

SetState sets the state for the Euler stepper.

func (*Euler) SetStep

func (euler *Euler) SetStep(step float64) error

SetStep sets the state for the Euler stepper.

func (*Euler) State

func (euler *Euler) State() []float64

State returns the state of the Euler stepper.

func (*Euler) Step

func (euler *Euler) Step() ([]float64, error)

Step performs one step iteration call of the Euler stepper. It also updates the state of the Euler object.

The Euler method for the system,

y = f(t, y)

Consists of building the sequence of numbers t_n, y_n, following the recurrence,

t_n+1 = t_n + dt
y_n+1 = y_n + dt*f(t_n, y_n) // First step

func (*Euler) StepSize

func (euler *Euler) StepSize() float64

StepSize returns the step size of the method

type Midpoint

type Midpoint struct {
	// contains filtered or unexported fields
}

Midpoint implements the Midpoint stepper method. Midpoint is part of the Stepper interface.

func NewMidpoint

func NewMidpoint(stepSize float64, system System) (r *Midpoint)

NewMidpoint returns a reference to a new Midpoint stepper method.

func (*Midpoint) Set

func (midpoint *Midpoint) Set(stepSize float64, system System) error

Set sets the step size and System for the Midpoint stepper.

func (*Midpoint) SetState

func (midpoint *Midpoint) SetState(state []float64) error

SetState sets the state for the Midpoint stepper.

func (*Midpoint) SetStep

func (midpoint *Midpoint) SetStep(step float64) error

SetStep sets the state for the Midpoint stepper.

func (*Midpoint) State

func (midpoint *Midpoint) State() []float64

State returns the state of the Midpoint stepper.

func (*Midpoint) Step

func (midpoint *Midpoint) Step() ([]float64, error)

Step performs one step iteration call of the Midpoint stepper. It also updates the state of the Midpoint object.

The mid-point method for the system,

y = f(t, y)

Consists of building the sequence of numbers t_n, y_n, following the recurrence,

t_n+1 = t_n + dt
ya_n = y_n + 0.5*dt*f(t_n, y_n) // First step
y_n+1 = y_n + dt*f(t_n, ya_n)   // Second step

func (*Midpoint) StepSize

func (midpoint *Midpoint) StepSize() float64

StepSize returns the step size of the method

type Rk4

type Rk4 struct {
	// contains filtered or unexported fields
}

Rk4 implements the rk4 stepper method. rk4 is part of the Stepper interface.

func NewRk4

func NewRk4(stepSize float64, system System) (r *Rk4)

NewRk4 returns a reference to a new rk4 stepper method.

func (*Rk4) Set

func (rk4 *Rk4) Set(stepSize float64, system System) error

Set sets the step size and System for the Rk4 stepper.

func (*Rk4) SetState

func (rk4 *Rk4) SetState(state []float64) error

SetState sets the state for the Rk4 stepper.

func (*Rk4) SetStep

func (rk4 *Rk4) SetStep(step float64) error

SetStep sets the state for the Rk4 stepper.

func (*Rk4) State

func (rk4 *Rk4) State() []float64

State returns the state of the Rk4 stepper.

func (*Rk4) Step

func (rk4 *Rk4) Step() ([]float64, error)

Step performs one step iteration call of the Rk4 stepper. It also updates the state of the Rk4 object.

The Runge-Kutta4 method for the system,

y = f(t,y)

Consists of building the sequence of numbers t_n, y_n, following the recurrence,

t_n+1 = t_n + dt
k1 = dt*f(t_n, y_n)
k2 = dt*f(t_n + 0.5*dt, y_n + 0.5*k1)
k3 = dt*f(t_n + 0.5*dt, y_n + 0.5*k2)
k4 = dt*f(t_n + dt, y_n + k3)
y_n+1 = y_n + (1/6)*(k1 + 2*k2 + 2*k3 + k4) // First step

func (*Rk4) StepSize

func (rk4 *Rk4) StepSize() float64

StepSize returns the step size of the method

type Stepper

type Stepper interface {
	// Setter methods
	SetStep(step float64) error
	SetState(state []float64) error
	Set(stepSize float64, system System) error

	// Getter methods
	StepSize() float64
	State() []float64
	Step() ([]float64, error)
}

Stepper defines the functions that any Ordinary Differential Equation Integrator Stepper should implement.

type System

type System struct {
	// contains filtered or unexported fields
}

System wraps the function that represents the right-hand side of the ordinary differential equations system, its parameters t, and state x Consider the system,

x'(t) = f(x,t)  (1)

System wraps around f(x,t) storing also t and x. stateVector is the present state of the system, i.e. the values stored at the components of x. parametersVector are the parameters of to the system (i.e. the t in f(x,t)) function describes the func(state []float64, parameters []float64) []float64 that represents the right hand side of system (1)

func NewSystem

func NewSystem(state []float64, parameters []float64, system func(state []float64, parameters []float64) []float64) (s *System)

NewSystem returns a reference to a new System object with the properties given

func (*System) Evaluate

func (s *System) Evaluate(state []float64) []float64

Evaluate returns the result of evaluating f(x,t) with x = state. if the size of state is zero, it returns f(x,t) using the internal state x

func (*System) Function

func (s *System) Function() func(state []float64, parameters []float64) []float64

State returns the internal function of the System

func (*System) Parameters

func (s *System) Parameters() []float64

Parameters returns the internal parameters vector of the System

func (*System) State

func (s *System) State() []float64

State returns the internal state vector of the System

Jump to

Keyboard shortcuts

? : This menu
/ : Search site
f or F : Jump to
y or Y : Canonical URL