Herman Sletmoen

Type-stable solving with non-numeric parameters?

11

I am solving ODEs that depend on non-numeric parameters. The concrete parameter type differs between problems (e.g. with/without AD), so I can only type it abstractly in the system definition...

Big slowdown of ODE with matrix equations

7

This "matrix ODE" is much slower and allocates much more than it used to: ```julia using ModelingToolkit, OrdinaryDiffEq using ModelingToolkit: t_nounits as t, D_nounits as D @variables x(t)[1:2,1:1] eqs =...

An attempt at fixing https://github.com/SciML/ModelingToolkit.jl/issues/3659. Is there a cleaner way to do it, without the `if change_origin` "hack"?

Cannot initialize non-1-indexed array variables

2

Trivial example: ```julia using ModelingToolkit using ModelingToolkit: t_nounits as t, D_nounits as D @variables x(t)[0:1] @named M = ODESystem([D(x[0]) ~ 0, x[1] ~ 1], t) M = structural_simplify(M) prob =...

Automatic differentiation no longer works for an ODE system with a spline parameter

1

```julia using ModelingToolkit, OrdinaryDiffEq, ForwardDiff, DataInterpolations using ModelingToolkit: t_nounits as t, D_nounits as D splval(spl::LinearInterpolation, t) = spl(t) @register_symbolic splval(spl::LinearInterpolation, t) vars = @variables x(t) pars = @parameters P spl::LinearInterpolation...

Parameter dependencies fail with array variables

1

**Bug 1:** This should work (it works if you turn `c` and `d` into scalars), but fails: ```julia using ModelingToolkit using ModelingToolkit: t_nounits as t, D_nounits as D @parameters a...

Customize variable Latex display

5

I propose this to (optionally) let users customize how variables are displayed in Latex. - Default behavior is the same as before (untouched single-letter variables, and `\mathtt`-ed multi-letter variables). -...

[The documentation mentions an inplace function for evaluating a solution](https://docs.sciml.ai/DiffEqDocs/stable/basics/solution/), but this does not work with ModelingToolkit: ```julia using OrdinaryDiffEq, ModelingToolkit, Test using ModelingToolkit: t_nounits as t, D_nounits as D...

I'm trying to fix https://github.com/JuliaSymbolics/Symbolics.jl/issues/1299. Following @ChrisRackauckas suggestion, I have tried to add the code in this PR. But it does not work yet: ```julia using SymbolicUtils simplify(7.0) # outputs...

MvNormal unnecessarily slow when mean and covariance matrix have different types

3

My model has a `MvNormal` with a constant covariance matrix (just `Float64`s) and a differentiable mean vector (e.g. with `ForwardDiff.Dual`s): ```julia using Distributions, ForwardDiff, PDMats, LinearAlgebra, BenchmarkTools Σ = PDMat(Matrix(1.0I,...