CCDF bar plot for model predictions

mc_obs_ccdf(...)

mc_model_ccdf(..., n_sample = NA, group_sample = "collapse")

Arguments

...: Augments passed to ggdist::stat_ccdfinterval.
n_sample: The number of sample of draws to show in CCDF bar plot. Default NA. If n_sample is NA, then mc_model_ccdf() will use all draws from posterior distribution.
group_sample: How to show draws? Default "collapse". If group_sample is "collapse", then all draws are collapsed together to show in one CCDF bar plot; if group_sample is "group", then each draw is shown in an individual CCDF bar plot; if group_sample is "hops", then mc_model_ccdf() will use animation to show each draw in one frame; if group_sample is an function, then all draws are aggregated by group_sample(). See examples for more details.

Examples

library(ggplot2)

mcplot(mpg_model) +
  mc_model_ccdf(n_sample = 50) +
  mc_obs_ccdf() +
  mc_gglayer(coord_flip())


mcplot(mpg_model) +
  mc_model_ccdf(n_sample = 50) +
  mc_obs_ccdf() +
  mc_condition_on(x = vars(vs))


mcplot(mpg_model) +
  mc_model_ccdf(n_sample = 50, group_sample = mean) +
  mc_obs_ccdf() +
  mc_condition_on(x = vars(vs))


mcplot(mpg_model) +
  mc_model_ccdf(n_sample = 50, group_sample = mean) +
  mc_observation_transformation(mean) +
  mc_obs_ccdf() +
  mc_condition_on(x = vars(vs))