I am using mode_compute = 6. But, I have the following message:
marginal density: I’m computing the posterior mean and covariance… Done!
marginal density: I’m computing the posterior log marginal density (modified harmonic mean)… marginal density: the covariance of MCMC draws is not positive definite. You may have too few MCMC draws.Done!
Actually, I figured it out after watching your YouTube video! Thanks a lot! @jpfeifer
I have one more question about the measurement error in Garcia-Cicco et al. (2010) AER paper.
I’m trying to derive 1/4 of the standard deviation of ( g_y ), ( g_i ), ( g_c ), and ( tb_y ) using the Excel file. This perfectly matches the max values of the Uniform distribution shown in Table 3 of the paper.
However, in your code, you apply sqrt and mention that these are actually variances, not standard deviations.
The standard deviations of the estimated measurement error reported in Table 3 of the paper are actually variances. This mod-file reflects this difference to the published version by taking the square root.
That is, the files used to generate the results do something else than what was written in the paper. My codes reflect the original files.
My statement is one about the measurement errors not the data standard deviations. Instead of setting the standard deviations of the measurement errors to the corresponding standard deviation of the data, they set the variance of the measurement errors to the corresponding standard deviation of the data.
