Some stuff to discuss before generating the final results.
[x] It has the comment Original RBC Julia scripts used --adapts_burnin_prop=0.1 instead of the current 0.2, so perfect replication requires adding this to all Julia estimation script calls (or directly setting it in the defaults). right now. Not sure I understand this. Lets make sure things are fully replicable using the scripts as is?
More generally, do we want to doublecheck that adapts_burnin_prop is standardized?
[x] Where do we need to set a single thread for julia experiments?
I modified the baseline, but we should doublecheck the results
[x] The RBC2 SV didn't have a provided --init_params_file. I added one, but we should doublecheck this is what we want (and repalce if we change parameters)
I think we want to set this globally to be consistent. I have changed it in the defaults for RBC. Dynare does the same thing.
[x] Some burnins are called burnin_dynare_ such as the robustness ones. may want to rename them. Not exactly from dynare, just trying to be consisetnt with dynare initial conditions.
Some stuff to discuss before generating the final results.
Original RBC Julia scripts used --adapts_burnin_prop=0.1 instead of the current 0.2, so perfect replication requires adding this to all Julia estimation script calls (or directly setting it in the defaults).right now. Not sure I understand this. Lets make sure things are fully replicable using the scripts as is?adapts_burnin_propis standardized?Omega_1is set manually in https://github.com/HighDimensionalEconLab/HMCExamples.jl/blob/main/scripts/sgu_replication/run.sh#L8-L12 for the SGU experiments--perturb_covariancefrom the default for RBC. eg. https://github.com/HighDimensionalEconLab/HMCExamples.jl/blob/main/scripts/julia_replication/robustness.sh#L8burnin_dynare_such as the robustness ones. may want to rename them. Not exactly from dynare, just trying to be consisetnt with dynare initial conditions.