# Differentiable @unpack-ing of function arguments

**URL:** <https://discourse.julialang.org/t/differentiable-unpack-ing-of-function-arguments/43146>\
**Category:** Numerics\
**Tags:** diffeq\
**Created:** [July 15, 2020, 7:52pm UTC](https://discourse.julialang.org/t/differentiable-unpack-ing-of-function-arguments/43146 "2020-07-15T19:52:02Z")\
**Posts on this page:** 10\
**Page:** 1

<div class="post-metadata">

**Author:** ![mefob](https://avatars.discourse-cdn.com/v4/letter/m/ea5d25/32.png) [@mefob](https://discourse.julialang.org/u/mefob)\
**Post date:** [July 15, 2020, 7:52pm UTC](https://discourse.julialang.org/t/differentiable-unpack-ing-of-function-arguments/43146/1 "2020-07-15T19:52:02Z")

</div>

What is the recommended approach to @unpack parameters in a differential equation when using gradient wrt these parameters:  
I could not get named tuples or dicts to work, this is what I tried:

```julia
    using DiffEqSensitivity
    using DifferentialEquations
    using Flux
    using Parameters    
    #p = [1.0, 2.0] # works, but can't @unpack this array
    p = (a=1.0, b=2.0) # named tuple: ERROR: MethodError: no method matching similar(::NamedTuple{(:a, :b),Tuple{Float64,Float64}})
    #p = Dict(:a=>1.0, :b=>2.0) # dictionary: ERROR: MethodError: no method matching similar(::Dict{Symbol,Float64})
    function f!(du, u, p, t)
        #a, b = p # works, but @unpack would be prettier
        @unpack a, b = p # causes gradient to fail with errors printed above
        @. du = a * u + b
    end
    @show gradient(
        () -> sum(solve(
                ODEProblem(f!, [0.0], (0.0, 10.0), p),
                Tsit5(), sensealg = QuadratureAdjoint())),        
        params(p))

```

---

<div class="post-metadata">

**Author:** ![ChrisRackauckas](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/chrisrackauckas/32/77_2.png) [@ChrisRackauckas](https://discourse.julialang.org/u/ChrisRackauckas)\
**Post date:** [July 15, 2020, 9:03pm UTC](https://discourse.julialang.org/t/differentiable-unpack-ing-of-function-arguments/43146/2 "2020-07-15T21:03:26Z")

</div>

> [@mefob](#):
>
> `a, b = p `

`a, b = p` (without the `@unpack`) should be fine.

---

<div class="post-metadata">

**Author:** ![mefob](https://avatars.discourse-cdn.com/v4/letter/m/ea5d25/32.png) [@mefob](https://discourse.julialang.org/u/mefob)\
**Post date:** [July 15, 2020, 9:23pm UTC](https://discourse.julialang.org/t/differentiable-unpack-ing-of-function-arguments/43146/3 "2020-07-15T21:23:41Z")

</div>

a, b = p works perfectly, but then I have to keep the arguments in the correct order. I am trying to keep track of about 100 different arguments, `@unpack` would be cleaner.

---

<div class="post-metadata">

**Author:** ![ChrisRackauckas](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/chrisrackauckas/32/77_2.png) [@ChrisRackauckas](https://discourse.julialang.org/u/ChrisRackauckas)\
**Post date:** [July 15, 2020, 9:28pm UTC](https://discourse.julialang.org/t/differentiable-unpack-ing-of-function-arguments/43146/4 "2020-07-15T21:28:21Z")

</div>

The difficulty is that you need something that is integrable. @jonniedie does `@unpack` work with ComponentArrays? That might be the solution here since the adjoints work.

---

<div class="post-metadata">

**Author:** ![jonniedie](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/jonniedie/32/12842_2.png) [@jonniedie](https://discourse.julialang.org/u/jonniedie)\
**Post date:** [July 15, 2020, 9:53pm UTC](https://discourse.julialang.org/t/differentiable-unpack-ing-of-function-arguments/43146/5 "2020-07-15T21:53:57Z")

</div>

`@unpack` works. It looks like I had a `convert` method missing for Zygote though (which, oddly enough, I’ve never run into in similar problems). The fix is getting registered now.

But I’m getting zero for the gradients. Not sure what’s missing:

```julia
using ComponentArrays
using DiffEqSensitivity
using DifferentialEquations
using Flux
using Parameters

p = ComponentArray(a=1.0, b=2.0)

function f!(du, u, p, t)
    # a, b = p # works, but @unpack would be prettier
    @unpack a, b = p
    @. du = a * u + b
end

@show grad = Flux.gradient(
        (par) -> sum(solve(
                ODEProblem(f!, [0.0], (0.0, 10.0), par),
                Tsit5(),
                sensealg = QuadratureAdjoint(autojacvec=ZygoteVJP()),
            )),        
        p)

```

Things seem to be working fine with Zygote:

```julia
julia> Zygote.gradient(x-> sum(convert(typeof(x), x).^2), p)
((a = 2.0, b = 4.0),)

julia> function f(u, p, t)
           @unpack a, b = p
           a .* u .+ b
       end
f (generic function with 1 method)

julia> gradient(x->sum(f([0.0], x, 0.0)), p)
((a = 0.0, b = 1.0),)

```

---

<div class="post-metadata">

**Author:** ![jonniedie](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/jonniedie/32/12842_2.png) [@jonniedie](https://discourse.julialang.org/u/jonniedie)\
**Post date:** [July 15, 2020, 10:05pm UTC](https://discourse.julialang.org/t/differentiable-unpack-ing-of-function-arguments/43146/6 "2020-07-15T22:05:23Z")

</div>

Looks like out-of-place works fine, though.

```julia
julia> @show Flux.gradient((par->begin
                #= C:\Users\jdiegelm\.julia\dev\ComponentArrays\examples\wip\diff.jl:21 =#
                sum(solve(ODEProblem(f, [0.0], (0.0, 10.0), par), Tsit5(), sensealg = QuadratureAdjoint(autojacvec = ZygoteVJP())))
            end), p) = ((a = 682437.0287268654, b = 40552.658586328536),)
((a = 682437.0287268654, b = 40552.658586328536),)

```

---

<div class="post-metadata">

**Author:** ![mefob](https://avatars.discourse-cdn.com/v4/letter/m/ea5d25/32.png) [@mefob](https://discourse.julialang.org/u/mefob)\
**Post date:** [July 15, 2020, 10:05pm UTC](https://discourse.julialang.org/t/differentiable-unpack-ing-of-function-arguments/43146/7 "2020-07-15T22:05:46Z")

</div>

`@unpack` works for me as well, thanks! I’ll check out your fix later.  
When I use `QuadratureAdjoint` I am running into `ERROR: type TrackedArray has no field a`:  
`getproperty(::ReverseDiff.TrackedArray{Int64,Float64,1,ComponentArray{Int64,1,Array{Int64,1},Tuple{Axis{(a = 1, b = 2)}}},ComponentArray{Float64,1,Array{Float64,1},Tuple{Axis{(a = 1, b = 2)}}}}, ::Symbol) at .\Base.jl:33`

```julia
       p = ComponentArray(a=1, b=2)
    function f!(du, u, p, t)
        @unpack a,b = p
        @. du = a * u + b
    end
    @show gradient(
        () -> sum(solve(
                ODEProblem(f!, [0.0], (0.0, 10.0), p),
                Tsit5(), sensealg = QuadratureAdjoint())),        
        params(p))

```

---

<div class="post-metadata">

**Author:** ![jonniedie](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/jonniedie/32/12842_2.png) [@jonniedie](https://discourse.julialang.org/u/jonniedie)\
**Post date:** [July 15, 2020, 10:08pm UTC](https://discourse.julialang.org/t/differentiable-unpack-ing-of-function-arguments/43146/8 "2020-07-15T22:08:38Z")

</div>

Yeah, it’s gotta be `QuadratureAdjoint(autojacvec=ZygoteVJP())`. Tracker and ReverseDiff don’t work with ComponentArrays

---

<div class="post-metadata">

**Author:** ![mefob](https://avatars.discourse-cdn.com/v4/letter/m/ea5d25/32.png) [@mefob](https://discourse.julialang.org/u/mefob)\
**Post date:** [July 15, 2020, 10:09pm UTC](https://discourse.julialang.org/t/differentiable-unpack-ing-of-function-arguments/43146/9 "2020-07-15T22:09:45Z")

</div>

ok, thanks for the explanation!

---

<div class="post-metadata">

**Author:** ![ChrisRackauckas](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/chrisrackauckas/32/77_2.png) [@ChrisRackauckas](https://discourse.julialang.org/u/ChrisRackauckas)\
**Post date:** [July 15, 2020, 10:27pm UTC](https://discourse.julialang.org/t/differentiable-unpack-ing-of-function-arguments/43146/10 "2020-07-15T22:27:47Z")

</div>

> [@jonniedie](#):
>
> Looks like out-of-place works fine, though.
> 
> ```julia
> 
> ```

In-place Zygote has never really worked for us. Technically buffers should work, but in reality it doesn’t so we always default away from it.