Plot posterior predictive trajectory — plot

Plot as a probability mass function (suffix _pmf) or as a fanchart (stacked confidence interval, suffix _fanchart).

plot_post_traj_* plots the posterior predictive trajectory only
plot_ppc_traj_* overlays the observed trajectory to the posterior predictive trajectory

Usage

plot_post_traj_pmf(
  obj,
  id,
  patient_id,
  max_score = NA,
  support = NULL,
  max_scale = NA
)

plot_post_traj_fanchart(
  obj,
  id,
  patient_id,
  max_score = NA,
  interval = c("eti", "hdi"),
  CI_level = seq(0.1, 0.9, 0.1),
  ...
)

plot_ppc_traj_pmf(obj, train, test, patient_id, max_score = NA, ...)

plot_ppc_traj_fanchart(obj, train, test, patient_id, max_score = NA, ...)

Arguments

obj: Stanfit object or matrix of replications, with rows corresponding to samples and columns corresponding to variables (there should be nrow(id) columns).
id: Dataframe linking index in obj to (Patient, Time) pairs, cf. output from get_index().
patient_id: Patient ID.
max_score: (Optional) Maximum value that the score can take. In plot_*_traj_pmf this will set support if it is not supplied. In plot_*_traj_fanchart this will set the y axis range. In plot_ppc_traj_* this will check the content of train and test.
support: Values that the discrete distribution can take. Can be NULL, in that case the support of the pmf is estimated from the data (cf. HuraultMisc::extract_distribution()).
max_scale: Maximum value that the legend display. If NA, this chosen automatically.
interval: Type of confidence of interval to display, one of "eti" for equal-tailed intervals and "hdi" for highest density interval.
CI_level: Vector of confidence level to plot for the fanchart.
...: arguments to pass to plot_post_traj_*. For plot_post_traj_fanchart(), arguments to pass to add_fanchart().
train: Training dataset used to obtain the fit.
test: Testing dataset used to obtain the fit (can be NULL).

Value

Ggplot