# Some problem when optimizing the ODEProblem parameters

**URL:** <https://discourse.julialang.org/t/some-problem-when-optimizing-the-odeproblem-parameters/113746>\
**Category:** General Usage\
**Tags:** optimization, modelingtoolkit\
**Created:** [May 2, 2024, 2:40pm UTC](https://discourse.julialang.org/t/some-problem-when-optimizing-the-odeproblem-parameters/113746 "2024-05-02T14:40:16Z")\
**Posts on this page:** 1\
**Showing post:** 4

<div class="post-metadata">

**Author:** ![SteffenPL](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/steffenpl/32/206270_2.png) [@SteffenPL](https://discourse.julialang.org/u/SteffenPL)\
**Post date:** [May 3, 2024, 3:59am UTC](https://discourse.julialang.org/t/some-problem-when-optimizing-the-odeproblem-parameters/113746/4 "2024-05-03T03:59:07Z")

</div>

Yeah, I think there was work on deriving the sensitivity equations on a symbolic level, but as far as I know it is not done yet. See

> [@Direct Forward Sensitivity Analysis of ODEs in ModelingToolkit.jl for a Subset of Parameters](https://discourse.julialang.org/t/direct-forward-sensitivity-analysis-of-odes-in-modelingtoolkit-jl-for-a-subset-of-parameters/104903):
>
> Hello, I am working on a project that involves setting up and solving an ODE system. I’ve defined my ODE system using ModelingToolkit.jl and not as a function. My ODE system has many parameters, and I’m interested in performing a Direct Forward Sensitivity Analysis on a subset of these parameters, not all of them. Post-analysis, I intend to apply Bayesian inference to the results further. Here’s a simplified overview of what I am trying to do: using ModelingToolkit, Symbolics, DifferentialEqu…

You might already tried it, but you can get the adjoint equations with this package:

> **[Direct Adjoint Sensitivities of Differential Equations · SciMLSensitivity.jl](https://docs.sciml.ai/SciMLSensitivity/dev/manual/direct_adjoint_sensitivities/)**
>
> Documentation for SciMLSensitivity.jl.

(The adjoint ODE computes the derivative of a loss with respect to parameters. It’s mathematically kind of the ‘symbolic derivative’, which you can get to compute this gradient.)

---

_[View the full topic](https://discourse.julialang.org/t/some-problem-when-optimizing-the-odeproblem-parameters/113746)._