Syntax for empirical_bayes function with pumas model

Estimation
1

Hello,

I am trying the following code:

ebes = empirical_bayes(model, pop, params, FOCE)

I also tried with a single Subject and LaplaceI.

I am getting the following error:

ERROR: MethodError: no method matching empirical_bayes(::PumasModel{(tvtr_ratio = 1, tvCL = 1, tvQ = 1, tvVp = 1, tvVc = 1, tvbio_logit = 1, tvdur = 1, tvktr = 1, ωCL_BSV = 1, ωQ_BSV = 1, ωVp_BSV = 1, ωVc_BSV = 1, ξbio_BSV = 1, ωdur_BSV = 1, ωktr_BSV = 1, ω²CL_BOV = 1, ω²Q_BOV = 1, ω²Vp_BOV = 1, ω²Vc_BOV = 1, ξ²bio_BOV = 1, ω²dur_BOV = 1, ω²ktr_BOV = 1, σ_prop = 1), 28, (:Depot, :Transit1, :Transit2, :Transit3, :Transit4, :Transit5, :Central, :Peripheral), ParamSet{NamedTuple{(:tvtr_ratio, :tvCL, :tvQ, :tvVp, :tvVc, :tvbio_logit, :tvdur, :tvktr, :ωCL_BSV, :ωQ_BSV, :ωVp_BSV, :ωVc_BSV, :ξbio_BSV, :ωdur_BSV, :ωktr_BSV, :ω²CL_BOV, :ω²Q_BOV, :ω²Vp_BOV, :ω²Vc_BOV, :ξ²bio_BOV, :ω²dur_BOV, :ω²ktr_BOV, :σ_prop), Tuple{RealDomain{Float64, TransformVariables.Infinity{true}, Float64}, RealDomain{Float64, TransformVariables.Infinity{true}, Float64}, RealDomain{Float64, TransformVariables.Infinity{true}, Float64}, RealDomain{Float64, TransformVariables.Infinity{true}, Float64}, RealDomain{Float64, TransformVariables.Infinity{true}, Float64}, RealDomain{TransformVariables.Infinity{false}, TransformVariables.Infinity{true}, Float64}, RealDomain{Float64, TransformVariables.Infinity{true}, Float64}, RealDomain{Float64, TransformVariables.Infinity{true}, Float64}, RealDomain{Float64, TransformVariables.Infinity{true}, Float64}, RealDomain{Float64, TransformVariables.Infinity{true}, Float64}, RealDomain{Float64, TransformVariables.Infinity{true}, Float64}, RealDomain{Float64, TransformVariables.Infinity{true}, Float64}, RealDomain{Float64, TransformVariables.Infinity{true}, Float64}, RealDomain{Float64, TransformVariables.Infinity{true}, Float64}, RealDomain{Float64, TransformVariables.Infinity{true}, Float64}, RealDomain{Float64, TransformVariables.Infinity{true}, Float64}, RealDomain{Float64, TransformVariables.Infinity{true}, Float64}, RealDomain{Float64, TransformVariables.Infinity{true}, Float64}, RealDomain{Float64, TransformVariables.Infinity{true}, Float64}, RealDomain{Float64, TransformVariables.Infinity{true}, Float64}, RealDomain{Float64, TransformVariables.Infinity{true}, Float64}, RealDomain{Float64, TransformVariables.Infinity{true}, Float64}, RealDomain{Float64, TransformVariables.Infinity{true}, Float64}}}}, var"#5#15", Pumas.TimeDispatcher{var"#6#16", var"#7#17"}, Pumas.TimeDispatcher{var"#9#19", var"#10#20"}, var"#12#22", Pumas.LinearODE, var"#13#23", var"#14#24", ModelingToolkit.ODESystem}, ::Subject{NamedTuple{(:DV,), Tuple{Vector{Union{Missing, Float64}}}}, Pumas.ConstantInterpolationStructArray{Vector{Float64}, NamedTuple{(:period, :sequence, :seq_n, :formulation, :isT), Tuple{Vector{Int64}, Vector{String}, Vector{Int64}, Vector{String}, Vector{Int64}}}, Symbol}, Vector{Pumas.Event{Float64, Float64, Float64, Float64, Float64, Float64, Symbol}}, Vector{Float64}}, ::NamedTuple{(:tvCL, :tvQ, :tvVp, :tvVc, :tvbio_logit, :tvdur, :tvktr, :ωCL_BSV, :ωQ_BSV, :ωVp_BSV, :ωVc_BSV, :ξbio_BSV, :ωdur_BSV, :ωktr_BSV, :ω²CL_BOV, :ω²Q_BOV, :ω²Vp_BOV, :ω²Vc_BOV, :ξ²bio_BOV, :ω²dur_BOV, :ω²ktr_BOV, :σ_prop), NTuple{22, Float64}}, ::Type{FOCE})
Closest candidates are:
  empirical_bayes(::PumasModel, ::Pumas.AbstractSubject, ::NamedTuple, ::Union{FO, FOCE, LaplaceI}; diffeq_options) at /build/_work/PumasSystemImages/PumasSystemImages/julia_depot/packages/Pumas/Td3Jp/src/estimation/likelihoods.jl:5007
  empirical_bayes(::PumasModel, ::Pumas.AbstractSubject, ::NamedTuple, ::NaivePooled; diffeq_options) at /build/_work/PumasSystemImages/PumasSystemImages/julia_depot/packages/Pumas/Td3Jp/src/estimation/likelihoods.jl:5025
  empirical_bayes(::PumasModel, ::Pumas.AbstractSubject, ::NamedTuple, ::MAP; diffeq_options) at /build/_work/PumasSystemImages/PumasSystemImages/julia_depot/packages/Pumas/Td3Jp/src/estimation/likelihoods.jl:5037

Could I get some help with this?

Thank you.

2

Hello Donald,

It should be FOCE()/LaplaceI() instead of FOCE/LaplaceI. It also needs to be subject for this method so something like

ebes = [empirical_bayes(model, sub, params, FOCE()) for sub in pop]
1 Like