Explicitly ignore derivatives of argument checks

Project.toml

@@ -1,7 +1,7 @@
 name = "Distributions"
 uuid = "31c24e10-a181-5473-b8eb-7969acd0382f"
 authors = ["JuliaStats"]
-version = "0.25.41"
+version = "0.25.42"
 
 [deps]
 ChainRulesCore = "d360d2e6-b24c-11e9-a2a3-2a2ae2dbcce4"

src/cholesky/lkjcholesky.jl

@@ -41,10 +41,11 @@ end
 #  -----------------------------------------------------------------------------
 
 function LKJCholesky(d::Int, η::Real, _uplo::Union{Char,Symbol} = 'L'; check_args::Bool=true)
-    if check_args
-        d > 0 || throw(ArgumentError("matrix dimension must be positive"))
-        η > 0 || throw(ArgumentError("shape parameter must be positive"))
-    end
+    @check_args(
+        LKJCholesky,
+        (d, d > 0, "matrix dimension must be positive"),
+        (η, η > 0, "shape parameter must be positive"),
+    )
     logc0 = lkj_logc0(d, η)
     uplo = _char_uplo(_uplo)
     T = Base.promote_eltype(η, logc0)

src/edgeworth.jl

@@ -14,7 +14,7 @@ struct EdgeworthZ{D<:UnivariateDistribution} <: EdgeworthAbstract
     n::Float64
 
     function EdgeworthZ{D}(d::UnivariateDistribution, n::Real; check_args::Bool=true) where {D<:UnivariateDistribution}
-        check_args && @check_args(EdgeworthZ, n > zero(n))
+        @check_args EdgeworthZ (n, n > zero(n))
         new{D}(d, n)
     end
 end
@@ -80,7 +80,7 @@ struct EdgeworthSum{D<:UnivariateDistribution} <: EdgeworthAbstract
     dist::D
     n::Float64
     function EdgeworthSum{D}(d::UnivariateDistribution, n::Real; check_args::Bool=true) where {D<:UnivariateDistribution}
-        check_args && @check_args(EdgeworthSum, n > zero(n))
+        @check_args EdgeworthSum (n, n > zero(n))
         new{D}(d, n)
     end
 end
@@ -95,14 +95,15 @@ var(d::EdgeworthSum) = d.n*var(d.dist)
 struct EdgeworthMean{D<:UnivariateDistribution} <: EdgeworthAbstract
     dist::D
     n::Float64
-    function EdgeworthMean{D}(d::T, n::Real) where {D<:UnivariateDistribution,T<:UnivariateDistribution}
+    function EdgeworthMean{D}(d::UnivariateDistribution, n::Real; check_args::Bool=true) where {D<:UnivariateDistribution}
         # although n would usually be an integer, no methods are require this
-        n > zero(n) ||
-            error("n must be positive")
+        @check_args EdgeworthMean (n, n > zero(n), "n must be positive")
         new{D}(d, Float64(n))
     end
 end
-EdgeworthMean(d::UnivariateDistribution,n::Real) = EdgeworthMean{typeof(d)}(d,n)
+function EdgeworthMean(d::UnivariateDistribution, n::Real; check_args::Bool=true)
+    return EdgeworthMean{typeof(d)}(d, n; check_args=check_args)
+end
 
 mean(d::EdgeworthMean) = mean(d.dist)
 var(d::EdgeworthMean) = var(d.dist) / d.n

src/matrix/lkj.jl

@@ -33,10 +33,11 @@ end
 #  -----------------------------------------------------------------------------
 
 function LKJ(d::Integer, η::Real; check_args::Bool=true)
-    if check_args
-        d > 0 || throw(ArgumentError("Matrix dimension must be positive."))
-        η > 0 || throw(ArgumentError("Shape parameter must be positive."))
-    end
+    @check_args(
+        LKJ,
+        (d, d > 0, "matrix dimension must be positive."),
+        (η, η > 0, "shape parameter must be positive."),
+    )
     logc0 = lkj_logc0(d, η)
     T = Base.promote_eltype(η, logc0)
     LKJ{T, typeof(d)}(d, T(η), T(logc0))
@@ -75,10 +76,11 @@ insupport(d::LKJ, R::AbstractMatrix) = isreal(R) && size(R) == size(d) && isone(
 mean(d::LKJ) = Matrix{partype(d)}(I, dim(d), dim(d))
 
 function mode(d::LKJ; check_args::Bool=true)
-    p, η = params(d)
-    if check_args
-        η > 1 || throw(ArgumentError("mode is defined only when η > 1."))
-    end
+    @check_args(
+        LKJ,
+        @setup((_, η) = params(d)),
+        (η, η > 1, "mode is defined only when η > 1."),
+    )
     return mean(d)
 end
 

src/multivariate/dirichlet.jl

@@ -26,9 +26,10 @@ struct Dirichlet{T<:Real,Ts<:AbstractVector{T},S<:Real} <: ContinuousMultivariat
     lmnB::S
 
     function Dirichlet{T}(alpha::AbstractVector{T}; check_args::Bool=true) where T
-        if check_args && !all(x -> x > zero(x), alpha)
-            throw(ArgumentError("Dirichlet: alpha must be a positive vector."))
-        end
+        @check_args(
+            Dirichlet,
+            (alpha, all(x -> x > zero(x), alpha), "alpha must be a positive vector."),
+        )
         alpha0 = sum(alpha)
         lmnB = sum(loggamma, alpha) - loggamma(alpha0)
         new{T,typeof(alpha),typeof(lmnB)}(alpha, alpha0, lmnB)
@@ -39,7 +40,7 @@ function Dirichlet(alpha::AbstractVector{T}; check_args::Bool=true) where {T<:Re
     Dirichlet{T}(alpha; check_args=check_args)
 end
 function Dirichlet(d::Integer, alpha::Real; check_args::Bool=true)
-    check_args && @check_args(Dirichlet, d > zero(d) && alpha > zero(alpha))
+    @check_args Dirichlet (d, d > zero(d)) (alpha, alpha > zero(alpha))
     return Dirichlet{typeof(alpha)}(Fill(alpha, d); check_args=false)
 end
 

src/multivariate/multinomial.jl

@@ -26,22 +26,16 @@ struct Multinomial{T<:Real, TV<:AbstractVector{T}} <: DiscreteMultivariateDistri
 end
 
 function Multinomial(n::Integer, p::AbstractVector{T}; check_args::Bool=true) where {T<:Real}
-    if check_args
-        if n < 0
-            throw(ArgumentError("n must be a nonnegative integer."))
-        end
-        if !isprobvec(p)
-            throw(ArgumentError("p = $p is not a probability vector."))
-        end
-    end
+    @check_args(
+        Multinomial,
+        (n, n >= 0),
+        (p, isprobvec(p), "p is not a probability vector."),
+    )
     return Multinomial{T,typeof(p)}(n, p)
 end
 
 function Multinomial(n::Integer, k::Integer; check_args::Bool=true)
-    if check_args
-        @check_args(Multinomial, n >= 0)
-        @check_args(Multinomial, k >= 1)
-    end
+    @check_args Multinomial (n, n >= 0) (k, k >= 1)
     return Multinomial{Float64, Vector{Float64}}(round(Int, n), fill(1.0 / k, k))
 end
 

src/univariate/continuous/arcsine.jl

@@ -32,7 +32,7 @@ struct Arcsine{T<:Real} <: ContinuousUnivariateDistribution
 end
 
 function Arcsine(a::T, b::T; check_args::Bool=true) where {T <: Real}
-    check_args && @check_args(Arcsine, a < b)
+    @check_args Arcsine a < b
     return Arcsine{T}(a, b)
 end
 

src/univariate/continuous/beta.jl

@@ -33,14 +33,14 @@ struct Beta{T<:Real} <: ContinuousUnivariateDistribution
 end
 
 function Beta(α::T, β::T; check_args::Bool=true) where {T<:Real}
-    check_args && @check_args(Beta, α > zero(α) && β > zero(β))
+    @check_args Beta (α, α > zero(α)) (β, β > zero(β))
     return Beta{T}(α, β)
 end
 
 Beta(α::Real, β::Real; check_args::Bool=true) = Beta(promote(α, β)...; check_args=check_args)
 Beta(α::Integer, β::Integer; check_args::Bool=true) = Beta(float(α), float(β); check_args=check_args)
 function Beta(α::Real; check_args::Bool=true)
-    check_args && @check_args(Beta, α > zero(α))
+    @check_args Beta (α, α > zero(α))
     Beta(α, α; check_args=false)
 end
 Beta() = Beta{Float64}(1.0, 1.0)
@@ -66,10 +66,12 @@ params(d::Beta) = (d.α, d.β)
 mean(d::Beta) = ((α, β) = params(d); α / (α + β))
 
 function mode(d::Beta; check_args::Bool=true)
-    (α, β) = params(d)
-    if check_args
-        (α > 1 && β > 1) || error("mode is defined only when α > 1 and β > 1.")
-    end
+    α, β = params(d)
+    @check_args(
+        Beta,
+        (α, α > 1, "mode is defined only when α > 1."),
+        (β, β > 1, "mode is defined only when β > 1."),
+    )
     return (α - 1) / (α + β - 2)
 end
 

src/univariate/continuous/betaprime.jl

@@ -33,14 +33,14 @@ struct BetaPrime{T<:Real} <: ContinuousUnivariateDistribution
 end
 
 function BetaPrime(α::T, β::T; check_args::Bool=true) where {T<:Real}
-    check_args && @check_args(BetaPrime, α > zero(α) && β > zero(β))
+    @check_args BetaPrime (α, α > zero(α)) (β, β > zero(β))
     return BetaPrime{T}(α, β)
 end
 
 BetaPrime(α::Real, β::Real; check_args::Bool=true) = BetaPrime(promote(α, β)...; check_args=check_args)
 BetaPrime(α::Integer, β::Integer; check_args::Bool=true) = BetaPrime(float(α), float(β); check_args=check_args)
 function BetaPrime(α::Real; check_args::Bool=true)
-    check_args && @check_args(BetaPrime, α > zero(α))
+    @check_args BetaPrime (α, α > zero(α))
     BetaPrime(α, α; check_args=false)
 end
 BetaPrime() = BetaPrime{Float64}(1.0, 1.0)

src/univariate/continuous/biweight.jl

@@ -8,7 +8,7 @@ struct Biweight{T<:Real} <: ContinuousUnivariateDistribution
 end
 
 function Biweight(μ::T, σ::T; check_args::Bool=true) where {T<:Real}
-    check_args && @check_args(Biweight, σ > zero(σ))
+    @check_args Biweight (σ, σ > zero(σ))
     return Biweight{T}(μ, σ)
 end
 

src/univariate/continuous/cauchy.jl

@@ -29,7 +29,7 @@ struct Cauchy{T<:Real} <: ContinuousUnivariateDistribution
 end
 
 function Cauchy(μ::T, σ::T; check_args::Bool=true) where {T<:Real}
-    check_args && @check_args(Cauchy, σ > zero(σ))
+    @check_args Cauchy (σ, σ > zero(σ))
     return Cauchy{T}(μ, σ)
 end
 

src/univariate/continuous/chi.jl

@@ -27,7 +27,7 @@ struct Chi{T<:Real} <: ContinuousUnivariateDistribution
 end
 
 function Chi(ν::Real; check_args::Bool=true)
-    check_args && @check_args(Chi, ν > zero(ν))
+    @check_args Chi (ν, ν > zero(ν))
     return Chi{typeof(ν)}(ν)
 end
 
@@ -71,9 +71,10 @@ entropy(d::Chi{T}) where {T<:Real} = (ν = d.ν;
 
 function mode(d::Chi; check_args::Bool=true)
     ν = d.ν
-    if check_args
-        ν >= 1 || error("Chi distribution has no mode when ν < 1")
-    end
+    @check_args(
+        Chi,
+        (ν, ν >= 1, "Chi distribution has no mode when ν < 1"),
+    )
     sqrt(ν - 1)
 end
 

src/univariate/continuous/chisq.jl

@@ -26,7 +26,7 @@ struct Chisq{T<:Real} <: ContinuousUnivariateDistribution
 end
 
 function Chisq(ν::Real; check_args::Bool=true)
-    check_args && @check_args(Chisq, ν > zero(ν))
+    @check_args Chisq (ν, ν > zero(ν))
     return Chisq{typeof(ν)}(ν)
 end
 

src/univariate/continuous/cosine.jl

@@ -14,7 +14,7 @@ struct Cosine{T<:Real} <: ContinuousUnivariateDistribution
 end
 
 function Cosine(μ::T, σ::T; check_args::Bool=true) where {T <: Real}
-    check_args && @check_args(Cosine, σ > zero(σ))
+    @check_args Cosine (σ, σ > zero(σ))
     return Cosine{T}(μ, σ)
 end
 

src/univariate/continuous/epanechnikov.jl

@@ -8,7 +8,7 @@ struct Epanechnikov{T<:Real} <: ContinuousUnivariateDistribution
 end
 
 function Epanechnikov(μ::T, σ::T; check_args::Bool=true) where {T<:Real}
-    check_args && @check_args(Epanechnikov, σ > zero(σ))
+    @check_args Epanechnikov (σ, σ > zero(σ))
     return Epanechnikov{T}(μ, σ)
 end
 

src/univariate/continuous/erlang.jl

@@ -21,12 +21,12 @@ struct Erlang{T<:Real} <: ContinuousUnivariateDistribution
 end
 
 function Erlang(α::Real, θ::Real; check_args::Bool=true)
-    check_args && @check_args(Erlang, isinteger(α) && α >= zero(α))
+    @check_args Erlang (α, isinteger(α)) (α, α >= zero(α))
     return Erlang{typeof(θ)}(α, θ)
 end
 
 function Erlang(α::Integer, θ::Real; check_args::Bool=true)
-    check_args && @check_args(Erlang, α >= zero(α))
+    @check_args Erlang (α, α >= zero(α))
     return Erlang{typeof(θ)}(α, θ)
 end
 
@@ -63,9 +63,10 @@ kurtosis(d::Erlang) = 6 / d.α
 
 function mode(d::Erlang; check_args::Bool=true)
     α, θ = params(d)
-    if check_args
-        α >= 1 || error("Erlang has no mode when α < 1")
-    end
+    @check_args(
+        Erlang,
+        (α, α >= 1, "Erlang has no mode when α < 1"),
+    )
     θ * (α - 1)
 end
 

src/univariate/continuous/exponential.jl

@@ -27,7 +27,7 @@ struct Exponential{T<:Real} <: ContinuousUnivariateDistribution
 end
 
 function Exponential(θ::Real; check_args::Bool=true)
-    check_args && @check_args(Exponential, θ > zero(θ))
+    @check_args Exponential (θ, θ > zero(θ))
     return Exponential{typeof(θ)}(θ)
 end
 

src/univariate/continuous/fdist.jl

@@ -27,7 +27,7 @@ struct FDist{T<:Real} <: ContinuousUnivariateDistribution
     ν2::T
 
     function FDist{T}(ν1::T, ν2::T; check_args::Bool=true) where T
-        check_args && @check_args(FDist, ν1 > zero(ν1) && ν2 > zero(ν2))
+        @check_args FDist (ν1, ν1 > zero(ν1)) (ν2, ν2 > zero(ν2))
         new{T}(ν1, ν2)
     end
 end

src/univariate/continuous/frechet.jl

@@ -30,7 +30,7 @@ struct Frechet{T<:Real} <: ContinuousUnivariateDistribution
 end
 
 function Frechet(α::T, θ::T; check_args::Bool=true) where {T <: Real}
-    check_args && @check_args(Frechet, α > zero(α) && θ > zero(θ))
+    @check_args Frechet (α, α > zero(α)) (θ, θ > zero(θ))
     return Frechet{T}(α, θ)
 end
 

src/univariate/continuous/gamma.jl

@@ -31,7 +31,7 @@ struct Gamma{T<:Real} <: ContinuousUnivariateDistribution
 end
 
 function Gamma(α::T, θ::T; check_args::Bool=true) where {T <: Real}
-    check_args && @check_args(Gamma, α > zero(α) && θ > zero(θ))
+    @check_args Gamma (α, α > zero(α)) (θ, θ > zero(θ))
     return Gamma{T}(α, θ)
 end
 

src/univariate/continuous/generalizedpareto.jl

@@ -38,7 +38,7 @@ struct GeneralizedPareto{T<:Real} <: ContinuousUnivariateDistribution
 end
 
 function GeneralizedPareto(μ::T, σ::T, ξ::T; check_args::Bool=true) where {T <: Real}
-    check_args && @check_args(GeneralizedPareto, σ > zero(σ))
+    @check_args GeneralizedPareto (σ, σ > zero(σ))
     return GeneralizedPareto{T}(μ, σ, ξ)
 end
 

src/univariate/continuous/gumbel.jl

@@ -29,7 +29,7 @@ struct Gumbel{T<:Real} <: ContinuousUnivariateDistribution
 end
 
 function Gumbel(μ::T, θ::T; check_args::Bool=true) where {T <: Real}
-    check_args && @check_args(Gumbel, θ > zero(θ))
+    @check_args Gumbel (θ, θ > zero(θ))
     return Gumbel{T}(μ, θ)
 end
 

src/univariate/continuous/inversegamma.jl

@@ -32,7 +32,7 @@ struct InverseGamma{T<:Real} <: ContinuousUnivariateDistribution
 end
 
 function InverseGamma(α::T, θ::T; check_args::Bool=true) where {T <: Real}
-    check_args && @check_args(InverseGamma, α > zero(α) && θ > zero(θ))
+    @check_args InverseGamma (α, α > zero(α)) (θ, θ > zero(θ))
     return InverseGamma{T}(α, θ)
 end
 

src/univariate/continuous/inversegaussian.jl

@@ -30,7 +30,7 @@ struct InverseGaussian{T<:Real} <: ContinuousUnivariateDistribution
 end
 
 function InverseGaussian(μ::T, λ::T; check_args::Bool=true) where {T<:Real}
-    check_args && @check_args(InverseGaussian, μ > zero(μ) && λ > zero(λ))
+    @check_args InverseGaussian (μ, μ > zero(μ)) (λ, λ > zero(λ))
     return InverseGaussian{T}(μ, λ)
 end
 

src/univariate/continuous/laplace.jl

@@ -29,7 +29,7 @@ struct Laplace{T<:Real} <: ContinuousUnivariateDistribution
 end
 
 function Laplace(μ::T, θ::T; check_args::Bool=true) where {T <: Real}
-    check_args && @check_args(Laplace, θ > zero(θ))
+    @check_args Laplace (θ, θ > zero(θ))
     return Laplace{T}(μ, θ)
 end
 

src/univariate/continuous/levy.jl

@@ -28,7 +28,7 @@ struct Levy{T<:Real} <: ContinuousUnivariateDistribution
 end
 
 function Levy(μ::T, σ::T; check_args::Bool=true) where {T<:Real}
-    check_args && @check_args(Levy, σ > zero(σ))
+    @check_args Levy (σ, σ > zero(σ))
     return Levy{T}(μ, σ)
 end