CI TRIGGER

ext/DiffEqNoiseProcessReverseDiffExt.jl

@@ -4,18 +4,18 @@ using DiffEqNoiseProcess, DiffEqBase, Random
 isdefined(Base, :get_extension) ? (import ReverseDiff) : (import ..ReverseDiff)
 
 @inline function DiffEqNoiseProcess.wiener_randn(rng::Random.AbstractRNG,
-    proto::ReverseDiff.TrackedArray)
+        proto::ReverseDiff.TrackedArray)
     ReverseDiff.track(convert.(eltype(proto.value), randn(rng, size(proto))))
 end
 @inline function DiffEqNoiseProcess.wiener_randn!(rng::AbstractRNG,
-    rand_vec::Array{<:ReverseDiff.TrackedReal
-    })
+        rand_vec::Array{<:ReverseDiff.TrackedReal
+        })
     rand_vec .= ReverseDiff.track.(randn.((rng,), typeof.(DiffEqBase.value.(rand_vec))))
 end
 @inline function DiffEqNoiseProcess.wiener_randn!(rng::AbstractRNG,
-    rand_vec::AbstractArray{
-        <:ReverseDiff.TrackedReal,
-    })
+        rand_vec::AbstractArray{
+            <:ReverseDiff.TrackedReal,
+        })
     rand_vec .= ReverseDiff.track.(randn.((rng,), typeof.(DiffEqBase.value.(rand_vec))))
 end
 

src/bridges.jl

@@ -51,7 +51,7 @@ function BrownianBridge!(t0, tend, W0, Wh, Z0 = nothing, Zh = nothing; kwargs...
 end
 
 function GeometricBrownianBridge(μ, σ, t0, tend, W0, Wend, Z0 = nothing, Zend = nothing;
-    kwargs...)
+        kwargs...)
     W = GeometricBrownianMotionProcess(μ, σ, t0, W0, Z0; kwargs...)
     h = tend - t0
     Wh = Wend - W0
@@ -66,7 +66,7 @@ function GeometricBrownianBridge(μ, σ, t0, tend, W0, Wend, Z0 = nothing, Zend
 end
 
 function GeometricBrownianBridge!(μ, σ, t0, tend, W0, Wh, Z0 = nothing, Zh = nothing;
-    kwargs...)
+        kwargs...)
     W = GeometricBrownianMotionProcess!(μ, σ, t0, W0, Z0; kwargs...)
     h = tend - t0
     Wh .-= W0

src/copy_noise_types.jl

@@ -53,7 +53,7 @@ function Base.copy(W::SimpleNoiseProcess)
 end
 
 function Base.copy(W::Union{NoiseWrapper, NoiseGrid, NoiseApproximation,
-    VirtualBrownianTree, BoxWedgeTail})
+        VirtualBrownianTree, BoxWedgeTail})
     Wnew = typeof(W)((getfield(W, x) for x in fieldnames(typeof(W)))...)
     copy!(Wnew, W)
 end

src/correlated_noisefunc.jl

@@ -26,7 +26,7 @@ CorrelatedWienerProcess!(Γ,t0,W0,Z0=nothing;kwargs...)
 where `Γ` is the constant covariance matrix.
 """
 function CorrelatedWienerProcess(Γ, t0, W0, Z0 = nothing;
-    rng = Xorshifts.Xoroshiro128Plus(rand(UInt64)))
+        rng = Xorshifts.Xoroshiro128Plus(rand(UInt64)))
     NoiseProcess{false}(t0, W0, Z0, construct_correlated_noisefunc(Γ)..., rswm = RSWM(),
         rng = rng)
 end
@@ -58,7 +58,7 @@ CorrelatedWienerProcess!(Γ,t0,W0,Z0=nothing;kwargs...)
 where `Γ` is the constant covariance matrix.
 """
 function CorrelatedWienerProcess!(Γ, t0, W0, Z0 = nothing;
-    rng = Xorshifts.Xoroshiro128Plus(rand(UInt64)))
+        rng = Xorshifts.Xoroshiro128Plus(rand(UInt64)))
     NoiseProcess{true}(t0, W0, Z0, construct_correlated_noisefunc!(Γ)...,
         rswm = RSWM(), rng = rng)
 end

src/geometric_bm.jl

@@ -11,7 +11,7 @@
         rand_val = wiener_randn(rng, typeof(dW))
     end
     new_val = @.. exp(drift * dt + X.σ * sqrt(dt) * rand_val)
-    return W[end] * (new_val - 1)
+    return W.u[end] * (new_val - 1)
 end
 
 #=
@@ -69,7 +69,7 @@
 end
 function (X::GeometricBrownianMotion!)(rand_vec, W, dt, u, p, t, rng) #dist!
     wiener_randn!(rng, rand_vec)
-    @.. rand_vec = W[end] * expm1(X.μ - (1 / 2) * X.σ * dt + X.σ * sqrt(dt) * rand_vec)
+    @.. rand_vec = W.u[end] * expm1(X.μ - (1 / 2) * X.σ * dt + X.σ * sqrt(dt) * rand_vec)
 end
 
 @doc doc"""

src/noise_interfaces/box_wedge_tail_interface.jl

@@ -308,7 +308,7 @@ function generate_boxes1(densf, Δr, Δa, Δz, rM, aM, offset = nothing, scale =
 end
 
 function generate_boxes2(densf, Δrmin, Δamin, Δzmin, Δrmax, Δamax, Δzmax, rM, aM,
-    offset = nothing)
+        offset = nothing)
     boxes = Array{typeof(Δrmin), 1}[]
     probability = Vector{typeof(Δrmin)}(undef, 0)
     # start with largest possible size, then subsequently decrease size and fill remaining space

src/noise_interfaces/common.jl

@@ -1,8 +1,8 @@
 DiffEqBase.has_reinit(i::AbstractNoiseProcess) = true
 function DiffEqBase.reinit!(W::Union{NoiseProcess, NoiseApproximation}, dt;
-    t0 = W.t[1],
-    erase_sol = true,
-    setup_next = false)
+        t0 = W.t[1],
+        erase_sol = true,
+        setup_next = false)
     if erase_sol
         resize!(W.t, 1)
         resize!(W.W, 1)
@@ -55,9 +55,9 @@ function DiffEqBase.reinit!(W::Union{NoiseProcess, NoiseApproximation}, dt;
 end
 
 function DiffEqBase.reinit!(W::VirtualBrownianTree, dt;
-    t0 = W.t[1],
-    erase_sol = false,
-    setup_next = true)
+        t0 = W.t[1],
+        erase_sol = false,
+        setup_next = true)
 
     # Back to noise's starting state
     W.curt = first(W.t)
@@ -88,9 +88,9 @@ function DiffEqBase.reinit!(W::VirtualBrownianTree, dt;
 end
 
 function DiffEqBase.reinit!(W::NoiseGrid, dt;
-    t0 = W.t[1],
-    erase_sol = true,
-    setup_next = false)
+        t0 = W.t[1],
+        erase_sol = true,
+        setup_next = false)
     W.curt = t0
     W.dt = dt
     if t0 == W.t[1]
@@ -119,9 +119,9 @@ function DiffEqBase.reinit!(W::NoiseGrid, dt;
 end
 
 function DiffEqBase.reinit!(W::AbstractNoiseProcess, dt;
-    t0 = W.t[1],
-    erase_sol = true,
-    setup_next = false)
+        t0 = W.t[1],
+        erase_sol = true,
+        setup_next = false)
     W.curt = t0
     W.dt = dt
 
@@ -130,9 +130,9 @@ function DiffEqBase.reinit!(W::AbstractNoiseProcess, dt;
 end
 
 function DiffEqBase.reinit!(W::NoiseFunction, dt;
-    t0 = W.t0,
-    erase_sol = true,
-    setup_next = false)
+        t0 = W.t0,
+        erase_sol = true,
+        setup_next = false)
     W.curt = t0
     W.dt = dt
 
@@ -156,9 +156,9 @@ function DiffEqBase.reinit!(W::NoiseFunction, dt;
 end
 
 function DiffEqBase.reinit!(W::NoiseTransport, dt;
-    t0 = W.t0,
-    erase_sol = true,
-    setup_next = false)
+        t0 = W.t0,
+        erase_sol = true,
+        setup_next = false)
     W.curt = t0
     W.dt = dt
 
@@ -188,9 +188,9 @@ function DiffEqBase.reinit!(W::NoiseTransport, dt;
 end
 
 function DiffEqBase.reinit!(W::NoiseWrapper, dt;
-    t0 = W.t[1],
-    erase_sol = true,
-    setup_next = false)
+        t0 = W.t[1],
+        erase_sol = true,
+        setup_next = false)
     W.curt = t0
     W.dt = dt
 

src/noise_interfaces/virtual_brownian_tree_interface.jl

@@ -162,7 +162,7 @@ end
 
 # create the cache of the VBT with depth tree_depth
 function create_VBT_cache(bridge, t0, W0, Z0, tend, Wend, Zend, rng::Random123.AbstractR123,
-    tree_depth, search_depth)
+        tree_depth, search_depth)
     # total number of cached time steps and W values
     Nt = Int(2^search_depth + 1)
 
@@ -230,7 +230,7 @@ function create_VBT_cache(bridge, t0, W0, Z0, tend, Wend, Zend, rng::Random123.A
 end
 
 function search_VBT(t, seed, t0, t1, W0, W1, Z0, Z1, W::VirtualBrownianTree,
-    rng::Random123.AbstractR123)
+        rng::Random123.AbstractR123)
     Nt = Int(2^W.search_depth + 1)
     depth = Int(W.tree_depth + 1)
     seed_l, seed_r, seed_v = split_VBT_seed(rng, seed, depth, Nt)
@@ -286,7 +286,7 @@ function search_VBT(t, seed, t0, t1, W0, W1, Z0, Z1, W::VirtualBrownianTree,
 end
 
 function search_VBT!(out1, out2, t, seed, t0, t1, W0, W1, Z0, Z1, W::VirtualBrownianTree,
-    rng::Random123.AbstractR123)
+        rng::Random123.AbstractR123)
     Nt = Int(2^W.search_depth + 1)
     depth = Int(W.tree_depth + 1)
     seed_l, seed_r, seed_v = split_VBT_seed(rng, seed, depth, Nt)

src/pCN.jl

@@ -24,7 +24,7 @@ External links
  * [Preconditioned Crank–Nicolson algorithm on Wikipedia](https://en.wikipedia.org/wiki/Preconditioned_Crank–Nicolson_algorithm)
 """
 function pCN!(source::AbstractNoiseProcess{T, N, Vector{T2}, inplace}, ρ;
-    reset = true, reverse = false, indx = nothing) where {T, N, T2, inplace}
+        reset = true, reverse = false, indx = nothing) where {T, N, T2, inplace}
 
     # generate new Wiener process similar to the one in source
     Wnew = generate_innovation(source.W[1], source.t, source.rng)
@@ -40,7 +40,7 @@ end
 Create a new, but correlated noise process from `noise` and additional entropy with correlation ρ.
 """
 function pCN(source::AbstractNoiseProcess{T, N, Vector{T2}, inplace}, ρ;
-    reset = true, reverse = false, indx = nothing) where {T, N, T2, inplace}
+        reset = true, reverse = false, indx = nothing) where {T, N, T2, inplace}
     source′ = deepcopy(source)
 
     # generate new Wiener process similar to the one in source
@@ -62,7 +62,7 @@ External links
   - [Preconditioned Crank–Nicolson algorithm on Wikipedia](https://en.wikipedia.org/wiki/Preconditioned_Crank%E2%80%93Nicolson_algorithm)
 """
 function pCN(source::NoiseGrid, ρ; reset = true,
-    rng = Xorshifts.Xoroshiro128Plus(rand(UInt64)))
+        rng = Xorshifts.Xoroshiro128Plus(rand(UInt64)))
 
     # generate new Wiener process similar to the one in source
     t = source.t

src/rswm.jl

@@ -4,8 +4,8 @@ mutable struct RSWM{T}
 end
 
 Base.@pure function RSWM(;
-    discard_length = 1e-15,
-    adaptivealg::Symbol = :RSwM3)
+        discard_length = 1e-15,
+        adaptivealg::Symbol = :RSwM3)
     RSWM{typeof(discard_length)}(discard_length, adaptivealg)
 end
 

src/solve.jl

@@ -1,6 +1,6 @@
 function DiffEqBase.__solve(prob::AbstractNoiseProblem,
-    args::Union{Nothing, SciMLBase.DEAlgorithm}...; dt = 0.0,
-    kwargs...)
+        args::Union{Nothing, SciMLBase.DEAlgorithm}...; dt = 0.0,
+        kwargs...)
     if dt == 0.0 || dt == nothing
         error("dt must be provided to simulate a noise process. Please pass dt=...")
     end