Skip to content

Fix Mooncake kwcall detection #130

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

Open
wants to merge 5 commits into
base: mooncake-ext
Choose a base branch
from
Open

Fix Mooncake kwcall detection #130

Changes from all commits
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
7b86a9a
Refine Mooncake rules
thorek1 Jul 22, 2025
074946b
Merge branch 'mooncake-ext' into fix-function-inlining-issue-in-moonc…
thorek1 Jul 22, 2025
97a9ec7
fix details on mooncake extension
Jul 22, 2025
d4db56b
fix mooncake ext
Jul 22, 2025
0ef5d7b
got get_NSSS_and_parameters to work. get_solution doesnt work, as he …
thorek1 Jul 25, 2025
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
296 changes: 275 additions & 21 deletions ext/MooncakeExt.jl
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4,46 +4,300 @@ import Mooncake
import MacroModelling
import SparseArrays
import LinearAlgebra as ℒ
import ChainRulesCore

Mooncake.@from_rrule Mooncake.DefaultCtx Tuple{typeof(MacroModelling.mul_reverse_AD!), Matrix{S}, AbstractMatrix{M}, AbstractMatrix{N}} where {S <: Real, M <: Real, N <: Real}
Mooncake.@from_rrule Mooncake.DefaultCtx Tuple{Any, typeof(MacroModelling.mul_reverse_AD!), Matrix{S}, AbstractMatrix{M}, AbstractMatrix{N}} where {S <: Real, M <: Real, N <: Real}

Mooncake.@from_rrule Mooncake.DefaultCtx Tuple{typeof(MacroModelling.sparse_preallocated!), Matrix{T}} where {T <: Real} true
Mooncake.@from_rrule Mooncake.DefaultCtx Tuple{Any, MacroModelling.higher_order_caches{S,F}, typeof(MacroModelling.sparse_preallocated!), Matrix{S}} where {S <: Real, F <: AbstractFloat} true

Mooncake.@from_rrule Mooncake.DefaultCtx Tuple{typeof(MacroModelling.calculate_second_order_stochastic_steady_state), Val{:newton}, Matrix{Float64}, SparseArrays.AbstractSparseMatrix{Float64}, Vector{Float64}, MacroModelling.ℳ} true
Mooncake.@from_rrule Mooncake.DefaultCtx Tuple{Any, AbstractFloat, typeof(MacroModelling.calculate_second_order_stochastic_steady_state), Val{:newton}, Matrix{Float64}, SparseArrays.AbstractSparseMatrix{Float64}, Vector{Float64}, MacroModelling.ℳ} true

Mooncake.@from_rrule Mooncake.DefaultCtx Tuple{typeof(MacroModelling.calculate_third_order_stochastic_steady_state), Val{:newton}, Matrix{Float64}, SparseArrays.AbstractSparseMatrix{Float64}, SparseArrays.AbstractSparseMatrix{Float64}, Vector{Float64}, MacroModelling.ℳ} true
Mooncake.@from_rrule Mooncake.DefaultCtx Tuple{Any, AbstractFloat, typeof(MacroModelling.calculate_third_order_stochastic_steady_state), Val{:newton}, Matrix{Float64}, SparseArrays.AbstractSparseMatrix{Float64}, SparseArrays.AbstractSparseMatrix{Float64}, Vector{Float64}, MacroModelling.ℳ} true

Mooncake.@from_rrule Mooncake.DefaultCtx Tuple{typeof(MacroModelling.calculate_jacobian), Vector{M}, Vector{N}, MacroModelling.ℳ} where {M <: AbstractFloat, N <: AbstractFloat}
# Mooncake.@from_rrule Mooncake.DefaultCtx Tuple{Any, MacroModelling.CalculationOptions, typeof(MacroModelling.get_NSSS_and_parameters), MacroModelling.ℳ, Vector{S}} where S <: AbstractFloat true

Mooncake.@from_rrule Mooncake.DefaultCtx Tuple{typeof(MacroModelling.calculate_hessian), Vector{M}, Vector{N}, MacroModelling.ℳ} where {M <: AbstractFloat, N <: AbstractFloat}
Mooncake.@from_rrule Mooncake.DefaultCtx Tuple{T¹, typeof(MacroModelling.get_NSSS_and_parameters), T², T³} where {T¹,T²,T³} true

Mooncake.@from_rrule Mooncake.DefaultCtx Tuple{typeof(MacroModelling.calculate_third_order_derivatives), Vector{M}, Vector{N}, MacroModelling.ℳ} where {M <: AbstractFloat, N <: AbstractFloat}
Mooncake.@from_rrule Mooncake.DefaultCtx Tuple{typeof(MacroModelling.get_NSSS_and_parameters), T¹,T²} where {T¹,T²} true

Mooncake.@from_rrule Mooncake.DefaultCtx Tuple{typeof(MacroModelling.get_NSSS_and_parameters), MacroModelling.ℳ, Vector{S} where S <: AbstractFloat} true
Mooncake.@from_rrule Mooncake.DefaultCtx Tuple{T¹, typeof(MacroModelling.calculate_jacobian), T², T³, T⁴} where {T¹,T²,T³,T⁴} true

Mooncake.@from_rrule Mooncake.DefaultCtx Tuple{typeof(MacroModelling.calculate_first_order_solution), Matrix{R}} where R <: AbstractFloat true
Mooncake.@from_rrule Mooncake.DefaultCtx Tuple{typeof(MacroModelling.calculate_jacobian), T², T³, T⁴} where {T²,T³,T⁴} true

Mooncake.@from_rrule Mooncake.DefaultCtx Tuple{typeof(MacroModelling.calculate_second_order_solution), AbstractMatrix{R}, SparseArrays.SparseMatrixCSC{R}, AbstractMatrix{R}, MacroModelling.second_order_auxiliary_matrices, MacroModelling.caches} where R <: AbstractFloat true
# Mooncake.@from_rrule Mooncake.DefaultCtx Tuple{Any, typeof(MacroModelling.calculate_jacobian), Vector{M}, Vector{N}, MacroModelling.ℳ} where {M <: AbstractFloat, N <: AbstractFloat}

Mooncake.@from_rrule Mooncake.DefaultCtx Tuple{typeof(MacroModelling.calculate_second_order_solution), AbstractMatrix{R}, SparseArrays.SparseMatrixCSC{R}, SparseArrays.SparseMatrixCSC{R}, AbstractMatrix{R}, SparseArrays.SparseMatrixCSC{R}, MacroModelling.second_order_auxiliary_matrices, MacroModelling.third_order_auxiliary_matrices, MacroModelling.caches} where R <: AbstractFloat true

Mooncake.@from_rrule Mooncake.DefaultCtx Tuple{typeof(MacroModelling.solve_lyapunov_equation), AbstractMatrix{R}, AbstractMatrix{R}} where R <: AbstractFloat true
Mooncake.@from_rrule Mooncake.DefaultCtx Tuple{typeof(MacroModelling.run_kalman_iterations), T³, T⁴, T⁵, T⁶,T⁷} where {T³,T⁴,T⁵,T⁶,T⁷} true

Mooncake.@from_rrule Mooncake.DefaultCtx Tuple{typeof(MacroModelling.solve_sylvester_equation), AbstractMatrix{R}, AbstractMatrix{R}, AbstractMatrix{R}} where R <: AbstractFloat true
Mooncake.@from_rrule Mooncake.DefaultCtx Tuple{T¹,T²,typeof(MacroModelling.run_kalman_iterations), T³, T⁴, T⁵, T⁶,T⁷} where {T¹,T²,T³,T⁴,T⁵,T⁶,T⁷} true

Mooncake.@from_rrule Mooncake.DefaultCtx Tuple{typeof(MacroModelling.find_shocks), Val{:LagrangeNewton}, Vector{Float64}, Vector{Float64}, AbstractMatrix{Float64}, ℒ.Diagonal{Bool, Vector{Bool}}, AbstractMatrix{Float64}, AbstractMatrix{Float64}, Vector{Float64}} true

Mooncake.@from_rrule Mooncake.DefaultCtx Tuple{typeof(MacroModelling.find_shocks), Val{:LagrangeNewton}, Vector{Float64}, Vector{Float64}, Vector{Float64}, AbstractMatrix{Float64}, AbstractMatrix{Float64}, AbstractMatrix{Float64},
ℒ.Diagonal{Bool, Vector{Bool}}, AbstractMatrix{Float64}, AbstractMatrix{Float64}, AbstractMatrix{Float64}, Vector{Float64}} true
Mooncake.@from_rrule Mooncake.DefaultCtx Tuple{Any, typeof(MacroModelling.calculate_hessian), Vector{M}, Vector{N}, MacroModelling.ℳ} where {M <: AbstractFloat, N <: AbstractFloat}

Mooncake.@from_rrule Mooncake.DefaultCtx Tuple{typeof(MacroModelling.calculate_inversion_filter_loglikelihood), Val{:first_order}, Vector{Vector{Float64}}, Matrix{Float64}, Matrix{Float64}, Union{Vector{String}, Vector{Symbol}}, MacroModelling.timings} true
Mooncake.@from_rrule Mooncake.DefaultCtx Tuple{Any, typeof(MacroModelling.calculate_third_order_derivatives), Vector{M}, Vector{N}, MacroModelling.ℳ} where {M <: AbstractFloat, N <: AbstractFloat}

Mooncake.@from_rrule Mooncake.DefaultCtx Tuple{typeof(MacroModelling.calculate_inversion_filter_loglikelihood), Val{:pruned_second_order},Vector{Vector{Float64}}, Vector{AbstractMatrix{Float64}}, Matrix{Float64}, Union{Vector{String}, Vector{Symbol}}, MacroModelling.timings} true
Mooncake.@from_rrule Mooncake.DefaultCtx Tuple{T¹, T², T³, T⁴, typeof(MacroModelling.calculate_first_order_solution), T⁵} where {T¹,T²,T³,T⁴,T⁵} true

Mooncake.@from_rrule Mooncake.DefaultCtx Tuple{typeof(MacroModelling.calculate_inversion_filter_loglikelihood), Val{:second_order},Vector{Float64}, Vector{AbstractMatrix{Float64}}, Matrix{Float64}, Union{Vector{String}, Vector{Symbol}}, MacroModelling.timings} true
Mooncake.@from_rrule Mooncake.DefaultCtx Tuple{typeof(MacroModelling.calculate_first_order_solution), T⁵} where {T⁵} true

Mooncake.@from_rrule Mooncake.DefaultCtx Tuple{typeof(MacroModelling.calculate_inversion_filter_loglikelihood), Val{:pruned_third_order},Vector{Vector{Float64}}, Vector{AbstractMatrix{Float64}}, Matrix{Float64}, Union{Vector{String}, Vector{Symbol}}, MacroModelling.timings} true
# Mooncake.@from_rrule Mooncake.DefaultCtx Tuple{Any, MacroModelling.timings, MacroModelling.CalculationOptions, AbstractMatrix{R}, typeof(MacroModelling.calculate_first_order_solution), Matrix{R}} where R <: AbstractFloat true

Mooncake.@from_rrule Mooncake.DefaultCtx Tuple{typeof(MacroModelling.calculate_inversion_filter_loglikelihood), Val{:third_order},Vector{Float64}, Vector{AbstractMatrix{Float64}}, Matrix{Float64}, Union{Vector{String}, Vector{Symbol}}, MacroModelling.timings} true
Mooncake.@from_rrule Mooncake.DefaultCtx Tuple{Any, MacroModelling.timings, AbstractMatrix{R}, MacroModelling.CalculationOptions, typeof(MacroModelling.calculate_second_order_solution), AbstractMatrix{R}, SparseArrays.SparseMatrixCSC{R}, AbstractMatrix{R}, MacroModelling.second_order_auxiliary_matrices, MacroModelling.caches} where R <: AbstractFloat true

Mooncake.@from_rrule Mooncake.DefaultCtx Tuple{Any, MacroModelling.timings, AbstractMatrix{R}, MacroModelling.CalculationOptions, typeof(MacroModelling.calculate_third_order_solution), AbstractMatrix{R}, SparseArrays.SparseMatrixCSC{R}, SparseArrays.SparseMatrixCSC{R}, AbstractMatrix{R}, SparseArrays.SparseMatrixCSC{R}, MacroModelling.second_order_auxiliary_matrices, MacroModelling.third_order_auxiliary_matrices, MacroModelling.caches} where R <: AbstractFloat true

Mooncake.@from_rrule Mooncake.DefaultCtx Tuple{Any, Int, Float64, typeof(MacroModelling.find_shocks), Val{:LagrangeNewton}, Vector{Float64}, Vector{Float64}, AbstractMatrix{Float64}, ℒ.Diagonal{Bool, Vector{Bool}}, AbstractMatrix{Float64}, AbstractMatrix{Float64}, Vector{Float64}} true

Mooncake.@from_rrule Mooncake.DefaultCtx Tuple{Any, Int, Float64, typeof(MacroModelling.find_shocks), Val{:LagrangeNewton}, Vector{Float64}, Vector{Float64}, Vector{Float64}, AbstractMatrix{Float64}, AbstractMatrix{Float64}, AbstractMatrix{Float64}, ℒ.Diagonal{Bool, Vector{Bool}}, AbstractMatrix{Float64}, AbstractMatrix{Float64}, AbstractMatrix{Float64}, Vector{Float64}} true

Mooncake.@from_rrule Mooncake.DefaultCtx Tuple{Any, Int, Int, MacroModelling.CalculationOptions, Symbol, typeof(MacroModelling.calculate_inversion_filter_loglikelihood), Val{:first_order}, Vector{Vector{Float64}}, Matrix{Float64}, Matrix{Float64}, Union{Vector{String}, Vector{Symbol}}, MacroModelling.timings} true

Mooncake.@from_rrule Mooncake.DefaultCtx Tuple{Any, Int, Int, MacroModelling.CalculationOptions, Symbol, typeof(MacroModelling.calculate_inversion_filter_loglikelihood), Val{:pruned_second_order}, Vector{Vector{Float64}}, Vector{AbstractMatrix{Float64}}, Matrix{Float64}, Union{Vector{String}, Vector{Symbol}}, MacroModelling.timings} true

Mooncake.@from_rrule Mooncake.DefaultCtx Tuple{Any, Int, Int, MacroModelling.CalculationOptions, Symbol, typeof(MacroModelling.calculate_inversion_filter_loglikelihood), Val{:second_order}, Vector{Float64}, Vector{AbstractMatrix{Float64}}, Matrix{Float64}, Union{Vector{String}, Vector{Symbol}}, MacroModelling.timings} true

Mooncake.@from_rrule Mooncake.DefaultCtx Tuple{Any, Int, Int, MacroModelling.CalculationOptions, Symbol, typeof(MacroModelling.calculate_inversion_filter_loglikelihood), Val{:pruned_third_order}, Vector{Vector{Float64}}, Vector{AbstractMatrix{Float64}}, Matrix{Float64}, Union{Vector{String}, Vector{Symbol}}, MacroModelling.timings} true

Mooncake.@from_rrule Mooncake.DefaultCtx Tuple{Any, Int, Int, MacroModelling.CalculationOptions, Symbol, typeof(MacroModelling.calculate_inversion_filter_loglikelihood), Val{:third_order}, Vector{Float64}, Vector{AbstractMatrix{Float64}}, Matrix{Float64}, Union{Vector{String}, Vector{Symbol}}, MacroModelling.timings} true

# Mooncake.@from_rrule Mooncake.DefaultCtx Tuple{Any, Symbol, AbstractFloat, AbstractFloat, Bool, typeof(MacroModelling.solve_lyapunov_equation), AbstractMatrix{R}, AbstractMatrix{R}} where R <: AbstractFloat true


Mooncake.@from_rrule Mooncake.DefaultCtx Tuple{T¹, T², T³, T⁴, typeof(MacroModelling.solve_lyapunov_equation), T⁵, T⁶} where {T¹,T²,T³,T⁴,T⁵,T⁶} true

Mooncake.@from_rrule Mooncake.DefaultCtx Tuple{typeof(MacroModelling.solve_lyapunov_equation), T⁵, T⁶} where {T⁵,T⁶} true


Mooncake.@from_rrule Mooncake.DefaultCtx Tuple{Any, AbstractMatrix{<:AbstractFloat}, Symbol, AbstractFloat, AbstractFloat, MacroModelling.sylvester_caches, Bool, typeof(MacroModelling.solve_sylvester_equation), AbstractMatrix{R}, AbstractMatrix{R}, AbstractMatrix{R}} where R <: AbstractFloat true


function ChainRulesCore.rrule(func_ir::Any,
ℂ::MacroModelling.higher_order_caches{T,F},
::typeof(MacroModelling.sparse_preallocated!),
Ŝ::Matrix{T}) where {T<:Real,F<:AbstractFloat}
ChainRulesCore.rrule(MacroModelling.sparse_preallocated!, Ŝ; ℂ=ℂ)
end

function ChainRulesCore.rrule(func_ir::Any,
::typeof(MacroModelling.mul_reverse_AD!),
C::Matrix{S},
A::AbstractMatrix{M},
B::AbstractMatrix{N}) where {S<:Real,M<:Real,N<:Real}
ChainRulesCore.rrule(MacroModelling.mul_reverse_AD!, C, A, B)
end

function ChainRulesCore.rrule(func_ir::Any,
::typeof(MacroModelling.calculate_jacobian),
parameters::Vector{M},
SS_and_pars::Vector{N},
m::MacroModelling.ℳ) where {M<:AbstractFloat,N<:AbstractFloat}
ChainRulesCore.rrule(MacroModelling.calculate_jacobian, parameters, SS_and_pars, m)
end

function ChainRulesCore.rrule(func_ir::Any,
::typeof(MacroModelling.calculate_hessian),
parameters::Vector{M},
SS_and_pars::Vector{N},
m::MacroModelling.ℳ) where {M<:AbstractFloat,N<:AbstractFloat}
ChainRulesCore.rrule(MacroModelling.calculate_hessian, parameters, SS_and_pars, m)
end

function ChainRulesCore.rrule(func_ir::Any,
::typeof(MacroModelling.calculate_third_order_derivatives),
parameters::Vector{M},
SS_and_pars::Vector{N},
m::MacroModelling.ℳ) where {M<:AbstractFloat,N<:AbstractFloat}
ChainRulesCore.rrule(MacroModelling.calculate_third_order_derivatives, parameters, SS_and_pars, m)
end

function ChainRulesCore.rrule(func_ir::Any,
tol::AbstractFloat,
::typeof(MacroModelling.calculate_second_order_stochastic_steady_state),
::Val{:newton},
S1::Matrix{Float64},
S2::SparseArrays.AbstractSparseMatrix{Float64},
x::Vector{Float64},
m::MacroModelling.ℳ)
ChainRulesCore.rrule(MacroModelling.calculate_second_order_stochastic_steady_state,
Val(:newton), S1, S2, x, m; tol=tol)
end

function ChainRulesCore.rrule(func_ir::Any,
tol::AbstractFloat,
::typeof(MacroModelling.calculate_third_order_stochastic_steady_state),
::Val{:newton},
S1::Matrix{Float64},
S2::SparseArrays.AbstractSparseMatrix{Float64},
S3::SparseArrays.AbstractSparseMatrix{Float64},
x::Vector{Float64},
m::MacroModelling.ℳ)
ChainRulesCore.rrule(MacroModelling.calculate_third_order_stochastic_steady_state,
Val(:newton), S1, S2, S3, x, m; tol=tol)
end

# function ChainRulesCore.rrule(func_ir::Any,
# opts::MacroModelling.CalculationOptions,
# ::typeof(MacroModelling.get_NSSS_and_parameters),
# m::MacroModelling.ℳ,
# x::Vector{S}) where {S<:AbstractFloat}
# ChainRulesCore.rrule(MacroModelling.get_NSSS_and_parameters, m, x; opts=opts)
# end

function ChainRulesCore.rrule(func_ir::Any,
T::MacroModelling.timings,
opts::MacroModelling.CalculationOptions,
initial_guess::AbstractMatrix{R},
::typeof(MacroModelling.calculate_first_order_solution),
∇₁::Matrix{R}) where {R<:AbstractFloat}
ChainRulesCore.rrule(MacroModelling.calculate_first_order_solution,
∇₁; T=T, opts=opts, initial_guess=initial_guess)
end

function ChainRulesCore.rrule(func_ir::Any,
T::MacroModelling.timings,
initial_guess::AbstractMatrix{R},
opts::MacroModelling.CalculationOptions,
::typeof(MacroModelling.calculate_second_order_solution),
∇₁::AbstractMatrix{R},
∇₂::SparseArrays.SparseMatrixCSC{R},
𝑺₁::AbstractMatrix{R},
M₂::MacroModelling.second_order_auxiliary_matrices,
ℂC::MacroModelling.caches) where {R<:AbstractFloat}
ChainRulesCore.rrule(MacroModelling.calculate_second_order_solution,
∇₁, ∇₂, 𝑺₁, M₂, ℂC; T=T, initial_guess=initial_guess, opts=opts)
end

function ChainRulesCore.rrule(func_ir::Any,
T::MacroModelling.timings,
initial_guess::AbstractMatrix{R},
opts::MacroModelling.CalculationOptions,
::typeof(MacroModelling.calculate_third_order_solution),
∇₁::AbstractMatrix{R},
∇₂::SparseArrays.SparseMatrixCSC{R},
∇₃::SparseArrays.SparseMatrixCSC{R},
𝑺₁::AbstractMatrix{R},
𝐒₂::SparseArrays.SparseMatrixCSC{R},
M₂::MacroModelling.second_order_auxiliary_matrices,
M₃::MacroModelling.third_order_auxiliary_matrices,
ℂC::MacroModelling.caches) where {R<:AbstractFloat}
ChainRulesCore.rrule(MacroModelling.calculate_third_order_solution,
∇₁, ∇₂, ∇₃, 𝑺₁, 𝐒₂, M₂, M₃, ℂC;
T=T, initial_guess=initial_guess, opts=opts)
end

function ChainRulesCore.rrule(func_ir::Any,
alg::Symbol,
tol::AbstractFloat,
acc_tol::AbstractFloat,
verbose::Bool,
::typeof(MacroModelling.solve_lyapunov_equation),
A::AbstractMatrix{R},
C::AbstractMatrix{R}) where {R<:AbstractFloat}
ChainRulesCore.rrule(MacroModelling.solve_lyapunov_equation,
A, C;
lyapunov_algorithm=alg,
tol=tol,
acceptance_tol=acc_tol,
verbose=verbose)
end

function ChainRulesCore.rrule(func_ir::Any,
initial_guess::AbstractMatrix{<:AbstractFloat},
syl_alg::Symbol,
acc_tol::AbstractFloat,
tol::AbstractFloat,
𝕊ℂ::MacroModelling.sylvester_caches,
verbose::Bool,
::typeof(MacroModelling.solve_sylvester_equation),
A::AbstractMatrix{R},
B::AbstractMatrix{R},
C::AbstractMatrix{R}) where {R<:AbstractFloat}
ChainRulesCore.rrule(MacroModelling.solve_sylvester_equation,
A, B, C;
initial_guess=initial_guess,
sylvester_algorithm=syl_alg,
acceptance_tol=acc_tol,
tol=tol,
𝕊ℂ=𝕊ℂ,
verbose=verbose)
end

function ChainRulesCore.rrule(func_ir::Any,
max_iter::Int,
tol::Float64,
::typeof(MacroModelling.find_shocks),
::Val{:LagrangeNewton},
initial_guess::Vector{Float64},
kron_buffer::Vector{Float64},
kron_buffer2::AbstractMatrix{Float64},
J::ℒ.Diagonal{Bool,Vector{Bool}},
S_i::AbstractMatrix{Float64},
S_i2e::AbstractMatrix{Float64},
shock_independent::Vector{Float64})
ChainRulesCore.rrule(MacroModelling.find_shocks,
Val(:LagrangeNewton),
initial_guess,
kron_buffer,
kron_buffer2,
J,
S_i,
S_i2e,
shock_independent;
max_iter=max_iter,
tol=tol)
end

function ChainRulesCore.rrule(func_ir::Any,
max_iter::Int,
tol::Float64,
::typeof(MacroModelling.find_shocks),
::Val{:LagrangeNewton},
initial_guess::Vector{Float64},
kron_buffer::Vector{Float64},
kron_buffer2::Vector{Float64},
kron_buffer3::AbstractMatrix{Float64},
kron_buffer4::AbstractMatrix{Float64},
kron_buffer5::AbstractMatrix{Float64},
J::ℒ.Diagonal{Bool,Vector{Bool}},
S_i::AbstractMatrix{Float64},
S_i2e::AbstractMatrix{Float64},
S_i3e::AbstractMatrix{Float64},
shock_independent::Vector{Float64})
ChainRulesCore.rrule(MacroModelling.find_shocks,
Val(:LagrangeNewton),
initial_guess,
kron_buffer,
kron_buffer2,
kron_buffer3,
kron_buffer4,
kron_buffer5,
J,
S_i,
S_i2e,
S_i3e,
shock_independent;
max_iter=max_iter,
tol=tol)
end

function ChainRulesCore.rrule(func_ir::Any,
warm_iters::Int,
presample::Int,
opts::MacroModelling.CalculationOptions,
filt_alg::Symbol,
::typeof(MacroModelling.calculate_inversion_filter_loglikelihood),
alg::Val{A},
state,
S,
data,
observables,
T::MacroModelling.timings) where A
ChainRulesCore.rrule(MacroModelling.calculate_inversion_filter_loglikelihood,
alg, state, S, data, observables, T;
warmup_iterations=warm_iters,
presample_periods=presample,
opts=opts,
filter_algorithm=filt_alg)
end

end # module
Loading