Fit beta distribution to MCMC output — fit_beta_prior • serosolver

Attempts to fit a beta distribution to a data frame of MCMC output from a previous run.

fit_beta_prior(
  chain_samples,
  par_name = "",
  error_tol = 999999999,
  try_attempts = 10,
  plot_fit = TRUE
)

Arguments

chain_samples: numeric vector of posterior samples to fit
par_name: optional parameter name used in the plot title
error_tol: tolerance for the optimisation error
try_attempts: maximum number of fitting attempts before giving up
plot_fit: if TRUE, plots the fitted distribution against the posterior density

Value

the model fit object returned by `optim`, or a list with missing parameters if fitting fails

See also

fit_normal_prior

Other priors: calc_phi_probs(), calc_phi_probs_indiv(), calc_phi_probs_spline(), create_prob_shifts(), dbb_prior(), density_beta_binom(), find_beta_prior_mode(), find_beta_prior_with_mean(), find_beta_prior_with_mean_var(), fit_normal_prior(), inf_mat_prior(), infection_history_prior(), prob_shifts()

Examples

if (FALSE) { # \dontrun{
## Posterior draws from a previous serosolver chain
chain <- rbeta(100, 2, 8)
results <- fit_beta_prior(chain, par_name="sigma1",plot_fit=FALSE)
} # }