Using mh_nblocks=4, as follows:
estimation(order=1, datafile='dsgedata.csv', mh_replic=20000, mh_nblocks=4, mh_jscale=2.) ;
The results appear to be for a single chain of length 20000:
Acceptance rate chain 1: 0.26595
Chains MCMC chain (20000×8×1 Array{Float64, 3}):
Iterations = 1:1:20000
Number of chains = 1
Samples per chain = 20000
In the context file, this seems to be confirmed:
julia> (context.results.model_results[1]).estimation.posterior_mcmc_chains_nbr
1
I can create an issue on Dynare.jl if that would be helpful.