-
The default choice of algorithm for ODEs tries to infer the desired level of precision by looking at number types: if it's not the "ordinary" Float32/64, it gives you a high-order algorithm because yo…
-
The following code to solve an ODE using DifferentialEquations.jl does not work in pyjulia. Strangely, the code works in a native Julia terminal (v1.0.2).
I am using the latest 'master' branch of p…
-
As far as I understand, Optimization.jl is focused on minimizing a function `f(u, p)` where `u` had to be an `AbstractArray` of some reasonable type (i.e. not array of arrays).
This set up can hand…
-
For some reason my program seem to be stuck at the very beginning (reading in a corpus, before training can be performed on that corpus) when I prepend the `train` function with `@trace`. For an hour …
-
Is there any way to use `LinearSolve.jl` algorithms with `RecursiveArrayTools.ArrayPartition`? Concrete example:
```julia
using RecursiveArrayTools
using DifferentialEquations
f!(du, u, p, t) = …
-
When I try to connect an equation of scalars with an equation of vectors, something on the compose step fails even though what is being connected is the same type (both scalar etc.). I am not sure why…
-
To end this saga of the Jacobian reuse changes which really came into fruition with https://github.com/JuliaDiffEq/OrdinaryDiffEq.jl/pull/927, we will want to do the following:
- [ ] OOP Wfact
- […
-
[`ComponentArrays.jl`](https://github.com/jonniedie/ComponentArrays.jl) is very helpful to handle states and parameters and works well with DifferentialEquations (e.g. plots are automatically labeled …
-
Thanks for the great package! I am trying to use BandedBlockBandedMatrix as a Jacobian in NLsolve (following https://github.com/JuliaDiffEq/DifferentialEquations.jl/issues/483). However the performanc…
-
`DiscreteBufferVariable`s should act like numbers instead of arrays. Instead of accessing them as `x[]`, just be able to do plain arithmetic on `x`. And we should access past values with `delay(x,1)` …