Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Make inference's job much easier by avoiding map #149

Merged
merged 5 commits into from
Aug 11, 2021
Merged

Make inference's job much easier by avoiding map #149

Changes from all commits
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
0d6f365
Make inference's job much easier by avoiding `map`
ChrisRackauckas Aug 11, 2021
e4cb322
A few more array specializations
ChrisRackauckas Aug 11, 2021
9f6d7f0
fix loop ordering
ChrisRackauckas Aug 11, 2021
ee7268c
fix tmp size
ChrisRackauckas Aug 11, 2021
73f2018
fix vect indexing
ChrisRackauckas Aug 11, 2021
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
67 changes: 56 additions & 11 deletions src/differentiation/compute_jacobian_ad.jl
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -14,20 +14,30 @@ void_setindex!(args...) = (setindex!(args...); return)

const default_chunk_size = ForwardDiff.pickchunksize

function ForwardColorJacCache(f,x,_chunksize = nothing;
function ForwardColorJacCache(f::F,x,_chunksize = nothing;
dx = nothing,
colorvec=1:length(x),
sparsity::Union{AbstractArray,Nothing}=nothing)
sparsity::Union{AbstractArray,Nothing}=nothing) where {F}

if _chunksize isa Nothing
chunksize = ForwardDiff.pickchunksize(maximum(colorvec))
else
chunksize = _chunksize
end

p = adapt.(parameterless_type(x),generate_chunked_partials(x,colorvec,chunksize))
_t = Dual{typeof(ForwardDiff.Tag(f,eltype(vec(x))))}.(vec(x),first(p))
t = ArrayInterface.restructure(x,_t)
if x isa Array
p = generate_chunked_partials(x,colorvec,chunksize)
t = similar(x,Dual{typeof(ForwardDiff.Tag(f,eltype(vec(x))))})
for i in eachindex(t)
t[i] = Dual{typeof(ForwardDiff.Tag(f,eltype(vec(x))))}(x[i],first(p)[1])
end
else
p = adapt.(parameterless_type(x),generate_chunked_partials(x,colorvec,chunksize))
_t = Dual{typeof(ForwardDiff.Tag(f,eltype(vec(x))))}.(vec(x),first(p))
t = ArrayInterface.restructure(x,_t)
end


if dx isa Nothing
fx = similar(t)
_dx = similar(x)
Expand All @@ -46,13 +56,27 @@ function generate_chunked_partials(x,colorvec,::Val{chunksize}) where chunksize
maxcolor = maximum(colorvec)
num_of_chunks = Int(ceil(maxcolor / chunksize))
padding_size = (chunksize - (maxcolor % chunksize)) % chunksize
partials = colorvec .== (1:maxcolor)'

# partials = colorvec .== (1:maxcolor)'
partials = BitMatrix(undef, length(colorvec), maxcolor)
for i in 1:maxcolor, j in 1:length(colorvec)
partials[j,i] = colorvec[j] == i
end

padding_matrix = BitMatrix(undef, length(x), padding_size)
partials = hcat(partials, padding_matrix)

chunked_partials = map(i -> Tuple.(eachrow(partials[:,(i-1)*chunksize+1:i*chunksize])),1:num_of_chunks)
chunked_partials

#chunked_partials = map(i -> Tuple.(eachrow(partials[:,(i-1)*chunksize+1:i*chunksize])),1:num_of_chunks)
chunked_partials = Vector{Vector{NTuple{chunksize,eltype(x)}}}(undef, num_of_chunks)
for i in 1:num_of_chunks
tmp = Vector{NTuple{chunksize,eltype(x)}}(undef, size(partials,1))
for j in 1:size(partials,1)
tmp[j] = Tuple(@view partials[j,(i-1)*chunksize+1:i*chunksize])
end
chunked_partials[i] = tmp
end
chunked_partials
end

@inline function forwarddiff_color_jacobian(f,
Expand Down Expand Up @@ -280,11 +304,26 @@ function forwarddiff_color_jacobian!(J::AbstractMatrix{<:Number},

for i in eachindex(p)
partial_i = p[i]
vect .= Dual{typeof(ForwardDiff.Tag(f,eltype(vecx)))}.(vecx, partial_i)

if vect isa Array
@inbounds @simd ivdep for j in eachindex(vect)
vect[j] = Dual{typeof(ForwardDiff.Tag(f,eltype(vecx)))}(vecx[j], partial_i[j])
end
else
vect .= Dual{typeof(ForwardDiff.Tag(f,eltype(vecx)))}.(vecx, partial_i)
end

f(fx,t)
if !(sparsity isa Nothing)
for j in 1:chunksize
dx .= partials.(fx, j)

if dx isa Array
@inbounds @simd ivdep for k in eachindex(dx)
dx[k] = partials(fx[k], j)
end
else
dx .= partials.(fx, j)
end

if ArrayInterface.fast_scalar_indexing(dx)
#dx is implicitly used in vecdx
Expand Down Expand Up @@ -313,7 +352,13 @@ function forwarddiff_color_jacobian!(J::AbstractMatrix{<:Number},
for j in 1:chunksize
col_index = (i-1)*chunksize + j
(col_index > ncols) && return J
J[:, col_index] .= partials.(vecfx, j)
if J isa Array
@inbounds @simd for k in 1:size(J,1)
J[k, col_index] = partials(vecfx[k], j)
end
else
J[:, col_index] .= partials.(vecfx, j)
end
end
end
end
Expand Down