Diagnostic of a set of COMPASS Models.
1 |
x |
a list of compass model fits of the same data with the same number of iterations, different seeds. Run some mcmc diagnostics on a series of COMPASS model fits. Assuming the input is a list of model fits for the same data with the same number of iterations and different seeds. Run Gelman's Rhat diagnostics on the alpha_s and alpha_u hyperparameter chains, treating each model as an independent chain. Rhat should be near 1 but rarely are in practice. Very large values may be a concern. The method returns an average model, by averaging the mean_gamma matrices (equally weighted since each input has the same number of iterations). This mean model should be better then any of the individual models. It can be plotted via "plot(result$mean_model)". |
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.