JuliaDiff / JuliaDiff/ReverseDiff.jl
Custom gradient for callable struct
Open
Nobody has claimed this yet.
- Dominant language
- Julia
- Stars
- 393
- Forks
- 60
- Avg merge
- 18h 24m
- Merged PRs (30d)
- 8
Description
This doesn't seem to work, simple example:
struct Over{T} den::T end
(o::Over)(x) = x ./ o.den
using ReverseDiff
(o::Over)(x::ReverseDiff.TrackedArray) = ReverseDiff.track(o, x)
ReverseDiff.@grad function (o::Over)(x)
ReverseDiff.value(x) ./ o.den, Δ -> (Δ .* o.den,)
end
# ERROR: UndefVarError: o not defined
This is the error after changing this line to closure = gensym(string(f)):
https://github.com/JuliaDiff/ReverseDiff.jl/blob/a7441cef4d5d1d8361946662bbab03856c2f6449/src/macros.jl#L182
or equivalently defining
Base.gensym(ex::Expr) = gensym(string(ex))
Contributor guide
No contributing guide indexed for this repository
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start with the callable-struct example in the issue and inspect the @grad macro around src/macros.jl line 182, especially how the callable expression is converted for gensym. Reproduce the UndefVarError, then verify that the custom gradient definition for Over works without the reported workaround.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- julia
- Domain
- tooling
- Issue type
- Bug
- Difficulty
- 3/5
- Estimated time
- 1-2 days
- Activity status
- Stale
- Clarity
- Mostly clear
- Newbie friendliness
- 42/100