Makes sense!
Here are the results for NaivePooled
FittedPumasModel
Successful minimization: true
Likelihood approximation: Pumas.NaivePooled
Objective function value: -8.5046991
Total number of observation records: 10
Number of active observation records: 10
Number of subjects: 2
Estimate
θ₁ 0.01
θ₂ 0.21883
θ₃ 0.049793
θ₄ 0.040644
θ₅ 0.052324
θ₆ 1.2279
σ_prop 0.01
Results for FOCEI are weirs specially theta_6
FittedPumasModel
Successful minimization: true
Likelihood approximation: Pumas.FOCEI
Objective function value: 18.396454
Total number of observation records: 10
Number of active observation records: 10
Number of subjects: 2
Estimate
θ₁ 0.020736
θ₂ 0.01
θ₃ 0.01
θ₄ 0.01
θ₅ 0.01
θ₆ 3.6819e17
σ_prop 0.36486
My constraints where for all theta as well as sigma_prop lower =0.01
NaivePooled with the 3 samples (res2)
gives
FittedPumasModel
Successful minimization: true
Likelihood approximation: Pumas.NaivePooled
Objective function value: -1.0753737
Total number of observation records: 15
Number of active observation records: 13
Number of subjects: 3
Estimate
θ₁ 0.01
θ₂ 6.4777
θ₃ 0.010113
θ₄ 0.10051
θ₅ 0.019268
θ₆ 0.010004
σ_prop 0.072009
#but neither
predict(res2)
#nor
infer(res2)
#nor
vpc(res2,10) |> plot
work
julia predict(res2)
returns
MethodError: no method matching _predict(::PumasModel{ParamSet{NamedTuple{(:θ, :σ_prop),Tuple{VectorDomain{Array{Float64,1},Array{TransformVariables.Infinity{true},1},Array{Float64,1}},RealDomain{Float64,TransformVariables.Infinity{true},Float64}}}},getfield(Main, Symbol("##353#360")),getfield(Main, Symbol("##354#361")),getfield(Main, Symbol("##355#362")),ODEProblem{Nothing,Tuple{Nothing,Nothing},false,Nothing,ODEFunction{false,getfield(Main, Symbol("##356#363")),UniformScaling{Bool},Nothing,Nothing,Nothing,Nothing,Nothing,Nothing,Nothing,Nothing,Nothing},Base.Iterators.Pairs{Union{},Union{},Tuple{},NamedTuple{,Tuple{}}},DiffEqBase.StandardODEProblem},getfield(Main, Symbol("##358#365")),getfield(Main, Symbol("##359#366"))}, ::Subject{NamedTuple{(:dv,),Tuple{Array{Union{Missing, Float64},1}}},Nothing,Array{Pumas.Event{Float64,Float64,Float64,Float64,Float64,Float64,Int64},1},Array{Float64,1}}, ::NamedTuple{(:θ, :σ_prop),Tuple{Array{Float64,1},Float64}}, ::Pumas.NaivePooled, ::Array{Any,1})
Closest candidates are:
_predict(::Any, ::Any, ::Any, !Matched::Pumas.FO, ::Any) at C:\Users\awolf-yadlin.juliapro\JuliaPro_v1.2.0-1\packages\Pumas\6uorK\src\estimation\diagnostics.jl:542
_predict(::Any, ::Any, ::Any, !Matched::Union{Pumas.FOCE, Laplace}, ::Any) at C:\Users\awolf-yadlin.juliapro\JuliaPro_v1.2.0-1\packages\Pumas\6uorK\src\estimation\diagnostics.jl:549
_predict(::Any, ::Any, ::Any, !Matched::Union{Pumas.FOCEI, Pumas.LaplaceI}, ::Any) at C:\Users\awolf-yadlin.juliapro\JuliaPro_v1.2.0-1\packages\Pumas\6uorK\src\estimation\diagnostics.jl:556
julia infer(res2)
returns
Calculating: variance-covariance matrix
MethodError: no method matching _orth_empirical_bayes!(::Array{Float64,1}, ::PumasModel{ParamSet{NamedTuple{(:θ, :σ_prop),Tuple{VectorDomain{Array{Float64,1},Array{TransformVariables.Infinity{true},1},Array{Float64,1}},RealDomain{Float64,TransformVariables.Infinity{true},Float64}}}},getfield(Main, Symbol("##353#360")),getfield(Main, Symbol("##354#361")),getfield(Main, Symbol("##355#362")),ODEProblem{Nothing,Tuple{Nothing,Nothing},false,Nothing,ODEFunction{false,getfield(Main, Symbol("##356#363")),UniformScaling{Bool},Nothing,Nothing,Nothing,Nothing,Nothing,Nothing,Nothing,Nothing,Nothing},Base.Iterators.Pairs{Union{},Union{},Tuple{},NamedTuple{,Tuple{}}},DiffEqBase.StandardODEProblem},getfield(Main, Symbol("##358#365")),getfield(Main, Symbol("##359#366"))}, ::Subject{NamedTuple{(:dv,),Tuple{Array{Union{Missing, Float64},1}}},Nothing,Array{Pumas.Event{Float64,Float64,Float64,Float64,Float64,Float64,Int64},1},Array{Float64,1}}, ::NamedTuple{(:θ, :σ_prop),Tuple{Array{Float64,1},Float64}}, ::Pumas.NaivePooled)
Closest candidates are:
_orth_empirical_bayes!(::AbstractArray{T,1} where T, ::PumasModel, ::Subject, ::NamedTuple, !Matched::Union{Pumas.FO, Pumas.FOI, Pumas.HCubeQuad}, !Matched::Any…; kwargs…) at C:\Users\awolf-yadlin.juliapro\JuliaPro_v1.2.0-1\packages\Pumas\6uorK\src\estimation\likelihoods.jl:228
_orth_empirical_bayes!(::AbstractArray{T,1} where T, ::PumasModel, ::Subject, ::NamedTuple, !Matched::Union{Pumas.FOCE, Pumas.FOCEI, Laplace, Pumas.LaplaceI}, !Matched::Any…; reltol, fdtype, fdrelstep, kwargs…) at C:\Users\awolf-yadlin.juliapro\JuliaPro_v1.2.0-1\packages\Pumas\6uorK\src\estimation\likelihoods.jl:245
and julia vpc(res2,10) |> plot
returns
MethodError: Cannot convert
an object of type Missing to an object of type Float64
Closest candidates are:
convert(::Type{T<:Real}, !Matched::Quantity) where T<:Real at C:\Users\awolf-yadlin.juliapro\JuliaPro_v1.2.0-1\packages\Unitful\ytsW0\src\conversion.jl:141
convert(::Type{T<:Real}, !Matched::Level) where T<:Real at C:\Users\awolf-yadlin.juliapro\JuliaPro_v1.2.0-1\packages\Unitful\ytsW0\src\logarithm.jl:22
convert(::Type{T<:Real}, !Matched::Gain) where T<:Real at C:\Users\awolf-yadlin.juliapro\JuliaPro_v1.2.0-1\packages\Unitful\ytsW0\src\logarithm.jl:62
which I assume ahve to do with the two missing values in the PKdata, but I am not sure how to deal with them in this context (ie how to remove them).
Moreover, given that infer doesn’t work I assume vpc will not work as it will lack the CIs for the shaded plots even if we didnt have the missing issue correct?
Thanks-
A
PS - I know my posts are long, but I am trying to provide as much info as possible!