search

HighDimensionalEconLab / HMCExamples.jl

MIT License
16 stars 3 forks source link

Verify number of MCMC draws for dynare robustness #158

Closed jlperla closed 1 year ago

jlperla commented 1 year ago

Looks like the code was doing for first order:

estimation(datafile = '../../data/rbc_1.csv', mode_file = 'rbc_mode.mat', order = 1, mh_nblocks = 1, mh_replic = 110000, mh_drop = 0.1, mh_jscale = 0.01, mode_compute = 0, mcmc_jumping_covariance = identity_matrix);

And for second order:

estimation(datafile = '../../data/rbc_2.csv', mode_file = 'rbc_mode.mat', order = 2, mh_nblocks = 1, mh_replic = 1250, mh_drop = 0.1, mode_compute = 0, mcmc_jumping_covariance = identity_matrix, number_of_particles = 20000, mh_jscale = 0.01);
donskerclass commented 1 year ago

Looks like the same as before, and no issues. Runtime may change relative to old figures due to execution variability, but current code will adjust labels for that.

Fine by me.