Individual level posterior distribution plots

General
,
1

Hi, I am very new to Stan. I have briefly used the hBayesDM package to estimate my model parameters and to plot the values at group and individual level. However, to over-run some bugs of the package I know need to use directly stan to estimate my parameters. I am wondering whether have some link/piece of script which would be useful in plotting individual posterior distributions. Its equivalent code in the hBayesDM was:

plotInd(output1, "ep")

Would stan_plot do it? or do I need to use bayes_plot package?
Thanks a lot for any piece of advice you could give!

2

The tidybayes package facilitates some great visualizations.

3

Thanks for your response. Do you know some equivalent code to the one provided above using the bayesplot package instead?

4

You can specify the parameters of interest and plot the distributions of those parameters using the various bayesplot functions.

For example:

mcmc_areas(stan_fit, pars = c("par1", "par2", "sigma"), prob = 0.89))
5

Hi,
I have a quick question on how to get the individual-level parameters posterior distributions in Stan.
I have this code running which gives me back the median value of the parameters across my subjects:

mcmc_areas(dataframe, pars = ("ind_alpha"), prob = 0.80, prob_outer = 0.95, point_est = c("median"))+
 labs(
   title = "Posterior distributions",
   subtitle = "with medians and 80% intervals"
 )

Whereas, I would want to plot the posterior distribution for each subject separately (and possibly still plotting the average value of the parameter). Do you know how could I do this?
Thanks a lot for your help and sorry if it is a very silly question.

6

Sorry for the very delayed reply!

If you extract samples of the “ind_alpha” parameters, you can then plot those directly.

With tidybayes it might look something like this:

spread_draws(Stan_Fit
            , ind_alpha[ID,]
            , ndraws = 250
             )