-
We are trying to write an `rrule` for a custom array class, such that `Zygote` can differentiate through it, but are stuck due to an error about a missing adjoint for a constructor. This may well be a…
-
With the release 1.6, I was reading a bit about precompilation. I noticed that Arrow package has quite high latency for the first time a file / message is read. I wonder if it would be practical to ge…
-
I've been thinking about about the approach for `WeightedParticles`. When I've needed this I haven't been able to use the built-in type. As it stands, the weighting vector is per-`Particles`, but I've…
-
It seems like all the links to `Point`, `Line` etc are broken in the documentation:
https://juliageo.org/GeoInterface.jl/stable/guides/defaults/#Wrapper-types
-
`@code_typed` (and `code_typed`) will sometimes report a different result than actually gets inferred if you call the function. For example:
```julia
julia> f1(x) = Base.typemax(x)
f1 (generic fu…
-
So extreme events analysis is currently very popular in nonlinear dynamics and we could get more functions that support this kind of stuff.
At the moment we have `mean_return_times` and `exit_entry…
-
Larger PDE solves may benefit from a reverse mode AD and also helps with adding support for neural networks as surrogates or part of training the PDEs themselves, therefore I was wondering what it wou…
-
Hi there,
Really nice package!
I was wondering if one can adjust the flatten/unflatten functions, such that unflatten is also working inside a closure for using Automatic Differentation.
A…
-
The Dask.distributed dynamic task scheduler could be replicated across different languages with low-to-moderate effort. This would require someone to build [Client](http://distributed.readthedocs.io/…
-
Would make the new version of Mocking which uses Cassette only work on Julia 0.7. I expect this work to start once all the major Cassette issues are resolved around the Julia 0.7 release.