ExponentialFamily.jl is a Julia package that extends the functionality of Distributions.jl by providing a collection of exponential family distributions and customized implementations.
ERROR: LoadError: Parameter vector RecursiveArrayTools.ArrayPartition{Float64, Tuple{Vector{Float64}, Matrix{Float64}}}(([2.681307820870243, 2.8111549974135714], [-1.7216965999028715 -0.4320207377364004; -0.43202073773640026 -2.325446169405964])) is not a valid natural parameter for distribution ExponentialFamily.MvNormalMeanCovariance.
The check verifies that the negative of [-1.7216965999028715 -0.4320207377364004; -0.43202073773640026 -2.325446169405964] is posdef. It is a correct check, but due to small differences in off diagonal entries the check fails. This is not ideal.
bvdmitri
commented
3 months ago
The check we have is this one
ERROR: LoadError: Parameter vector RecursiveArrayTools.ArrayPartition{Float64, Tuple{Vector{Float64}, Matrix{Float64}}}(([2.681307820870243, 2.8111549974135714], [-1.7216965999028715 -0.4320207377364004; -0.43202073773640026 -2.325446169405964])) is not a valid natural parameter for distribution ExponentialFamily.MvNormalMeanCovariance.The check verifies that the negative of
[-1.7216965999028715 -0.4320207377364004; -0.43202073773640026 -2.325446169405964]isposdef. It is a correct check, but due to small differences in off diagonal entries the check fails. This is not ideal.