It should work if you use
θ_guess ~ MvNormal(Fill(0.5, My*Nz), 0.01 * I)
Otherwise Turing does not sample θ_guess but only sets it to the distribution - which then leads to the errors.
1 Like
It should work if you use
θ_guess ~ MvNormal(Fill(0.5, My*Nz), 0.01 * I)
Otherwise Turing does not sample θ_guess but only sets it to the distribution - which then leads to the errors.