Overriding Base.getindex(A::OffsetArray{T,N}, I::Vararg{Int,N}) where {T,N} to print a stacktrace if it's in the second halo point,
function Base.getindex(A::OffsetArray{T,N}, I::Vararg{Int,N}) where {T,N}
any(I .< 0) && display(stacktrace())
@boundscheck checkbounds(A, I...)
J = map(parentindex, axes(A), I)
@inbounds parent(A)[J...]
end
I get three results which makes sense but can we rewrite so that the strain rate tensor doesn't dip into the second halo point?
getindex(::OffsetArray{Float64,3,Array{Float64,3}}, ::Int64, ::Int64, ::Int64) at REPL[74]:2
getindex at abstract_field.jl:172 [inlined]
ℑxᶠᵃᵃ at interpolation_operators.jl:8 [inlined]
U_over_ρ at compressible_operators.jl:15 [inlined]
∂xᶜᵃᵃ at derivative_operators.jl:18 [inlined]
strain_rate_tensor_ux at compressible_operators.jl:171 [inlined]
viscous_flux_ux at compressible_operators.jl:182 [inlined]
δxᶠᵃᵃ at difference_operators.jl:21 [inlined]
∂ⱼτ₁ⱼ at compressible_operators.jl:193 [inlined]
FU at right_hand_sides.jl:8 [inlined]
getindex(::OffsetArray{Float64,3,Array{Float64,3}}, ::Int64, ::Int64, ::Int64) at REPL[74]:2
getindex at abstract_field.jl:172 [inlined]
ℑyᵃᶠᵃ at interpolation_operators.jl:11 [inlined]
V_over_ρ at compressible_operators.jl:16 [inlined]
∂yᵃᶜᵃ at derivative_operators.jl:21 [inlined]
strain_rate_tensor_vy at compressible_operators.jl:172 [inlined]
viscous_flux_vy at compressible_operators.jl:183 [inlined]
δyᵃᶠᵃ at difference_operators.jl:24 [inlined]
∂ⱼτ₂ⱼ at compressible_operators.jl:199 [inlined]
FV at right_hand_sides.jl:13 [inlined]
getindex(::OffsetArray{Float64,3,Array{Float64,3}}, ::Int64, ::Int64, ::Int64) at REPL[74]:2
getindex at abstract_field.jl:172 [inlined]
ℑzᵃᵃᶠ at interpolation_operators.jl:14 [inlined]
W_over_ρ at compressible_operators.jl:17 [inlined]
∂zᵃᵃᶜ at derivative_operators.jl:24 [inlined]
strain_rate_tensor_wz at compressible_operators.jl:173 [inlined]
viscous_flux_wz at compressible_operators.jl:184 [inlined]
δzᵃᵃᶠ at difference_operators.jl:27 [inlined]
∂ⱼτ₃ⱼ at compressible_operators.jl:205 [inlined]
FW at right_hand_sides.jl:17 [inlined]
Overriding
Base.getindex(A::OffsetArray{T,N}, I::Vararg{Int,N}) where {T,N}
to print a stacktrace if it's in the second halo point,I get three results which makes sense but can we rewrite so that the strain rate tensor doesn't dip into the second halo point?