Type conversions for remaining Datasets

src/DataStructures/DataSetExact.jl

@@ -88,6 +88,13 @@ struct DataSetExact{XdistType<:Distribution, YdistType<:Distribution} <: Abstrac
 end
 
 
+function (::Type{T})(DSE::DataSetExact; kwargs...) where T<:Number
+	DataSetExact(ConvertDist(xdist(DSE),T), ConvertDist(ydist(DSE), T), dims(DSE), T.(yInvCov(DSE)),
+				(isnothing(DSE.WoundX) ? nothing : [SVector{xdim(DSE)}(Z) for Z in Windup(T.(xdata(DSE)), xdim(DSE))]); 
+                xnames=xnames(DSE), ynames=ynames(DSE), name=name(DSE), kwargs...)
+end
+
+
 # For SciMLBase.remake
 DataSetExact(;
 xdist::Distribution=Normal(0,1),
@@ -140,12 +147,6 @@ ynames(DSE::DataSetExact) = DSE.ynames
 name(DSE::DataSetExact) = DSE.name |> string
 
 
-function Base.BigFloat(DSE::DataSetExact; kwargs...)
-	BigDist(D::MultivariateNormal) = MvNormal(BigFloat.(mean(D)), BigFloat.(cov(D)))
-	remake(DSE; xdist=BigDist(xdist(DSE)), ydist=BigDist(ydist(DSE)), InvCov=BigFloat.(yInvCov(DSE)),
-				WoundX=(isnothing(DSE.WoundX) ? nothing : Windup(BigFloat.(xdata(DSE)), xdim(DSE))), kwargs...)
-end
-
 
 # function InformNames(DS::DataSetExact, xnames::AbstractVector{String}, ynames::AbstractVector{String})
 #     @assert length(xnames) == xdim(DS) && length(ynames) == ydim(DS)

src/DataStructures/DataSetUncertain.jl

@@ -65,6 +65,10 @@ struct DataSetUncertain <: AbstractUnknownUncertaintyDataSet
     end
 end
 
+function (::Type{T})(DS::DataSetUncertain; kwargs...) where T<:Number
+	DataSetUncertain(T.(xdata(DS)), T.(ydata(DS)), dims(DS), yinverrormodel(DS), SplitErrorParams(DS), T.(DS.testp), xnames(DS), ynames(DS), name(DS))
+end
+
 # For SciMLBase.remake
 DataSetUncertain(;
 x::AbstractVector=[0.],

src/DataStructures/DistributionTypes.jl

@@ -77,8 +77,24 @@ Distributions.cov(d::InformationGeometry.Dirac) = Diagonal(zeros(length(d)))
 Distributions.invcov(d::InformationGeometry.Dirac) = Diagonal([Inf for i in 1:length(d)])
 Distributions.pdf(d::InformationGeometry.Dirac, x::AbstractVector{<:Number}) = x == mean(d) ? 1.0 : 0.0
 Distributions.logpdf(d::InformationGeometry.Dirac, x::AbstractVector{<:Number}) = log(pdf(d, x))
+Distributions.params(d::InformationGeometry.Dirac) = (d.μ,)
 
 
 # Fix gradlogpdf for Cauchy distribution and product distributions in general
 Distributions.gradlogpdf(P::Cauchy,x::Real) = gradlogpdf(TDist(1), (x - P.μ) / P.σ) / P.σ
 Distributions.gradlogpdf(P::Product,x::AbstractVector) = [gradlogpdf(P.v[i],x[i]) for i in 1:length(x)]
+
+
+
+# Get Symbol for everything before {} in type.
+UnparametrizeType(D) = (S=string(typeof(D)); Symbol(S[1:findfirst('{',S)-1]))
+
+## Change Number Type of distributions
+ConvertDist(D::UnivariateDistribution, ::Type{T}) where T<:Number = eval(quote $(UnparametrizeType(D)){$T}($(T).($(params(D)))...) end)
+ConvertDist(D::InformationGeometry.Dirac, ::Type{T}) where T<:Number = InformationGeometry.Dirac(T.(D.μ))
+ConvertDist(D::MultivariateNormal, ::Type{T}) where T<:Number = MvNormal(T.(mean(D)), T.(cov(D)))
+
+function ConvertDist(D::Union{Distributions.Product, InformationGeometry.GeneralProduct}, ::Type{T}) where T<:Number
+    @assert eltype(D.v) <: Distribution{Univariate, Continuous}
+    product_distribution(broadcast(x->ConvertDist(x,T), D.v))
+end

src/DataStructures/GeneralizedDataSet.jl

@@ -54,6 +54,12 @@ struct GeneralizedDataSet{DistType<:Distribution} <: AbstractFixedUncertaintyDat
     end
 end
 
+function (::Type{T})(DS::GeneralizedDataSet; kwargs...) where T<:Number
+	GeneralizedDataSet(ConvertDist(dist(DS),T), dims(DS),
+				(isnothing(DS.WoundX) ? nothing : [SVector{xdim(DS)}(Z) for Z in Windup(T.(xdata(DS)), xdim(DS))]); 
+                xnames=xnames(DS), ynames=ynames(DS), name=name(DS), kwargs...)
+end
+
 # For SciMLBase.remake
 begin
     GeneralizedDataSet(;