Suppose I have the following function which takes the solution to an ODE and returns a sum-of-squared errors, making use of the integrator interface (which I need to use for my actual application, not necessary for this MWE): using DifferentialEquations, ForwardDiff f(u, p, t) = p * u u0 = 1 / 2 ts…

reinit! is designed to be non-allocating, allowing one to fill the values of the integrator and go again. You cannot fill an array of Float64’s with an array of Dual numbers, that’s just not possible because there isn’t even enough memory space there (dual numbers are larger bitsizes becaues they ar…

Updating integrator parameter/initial condition with ForwardDiff.Dual types

General Usage

DanielVandH July 6, 2022, 11:17am 4

Continued in Declaring ForwardDiff tag directly with a DifferentialEquations integrator / nested function

Topic		Replies	Views
Declaring ForwardDiff tag directly with a DifferentialEquations integrator / nested function General Usage optimization , forwarddiff , differentialequation	3	469	December 24, 2022
ForwardDiff and custom structure in DifferentialEquations New to Julia	2	751	October 24, 2020
Using ForwardDiff to differentiate an ODE with respect to some of its parameters Modelling & Simulations	5	615	April 10, 2020
How to update the initial condition in DifferentialEquations? General Usage diffeq	8	1419	January 9, 2020
Error with ForwardDiff no method matching Float64 Numerics	4	5712	June 23, 2020

Updating integrator parameter/initial condition with ForwardDiff.Dual types

Related topics