Plot the ECDF-based plots.

See vignette("rank_visualizations") for more details. See the methods for data_for_ecdf_plots() for available data formats.

plot_ecdf(
  x,
  variables = NULL,
  K = NULL,
  gamma = NULL,
  prob = 0.95,
  size = 1,
  alpha = 0.33,
  combine_variables = NULL,
  ecdf_alpha = NULL,
  ...,
  parameters = NULL
)

plot_ecdf_diff(
  x,
  variables = NULL,
  K = NULL,
  gamma = NULL,
  prob = 0.95,
  size = 1,
  alpha = 0.33,
  combine_variables = NULL,
  ecdf_alpha = NULL,
  ...,
  parameters = NULL
)

Arguments

x: object supporting the data_for_ecdf_plots() method.
variables: optional subset of variables to show in the plot
K: number of uniformly spaced evaluation points for the ECDF or ECDFs. Affects the granularity of the plot and can significantly speed up the computation of the simultaneous confidence bands. Defaults to the smaller of number of ranks per variable and the maximum rank.
gamma: TODO
prob: the width of the plotted confidence interval for the ECDF.
size: size passed to ggplot2::geom_ribbon() for the confidence band
alpha: alpha level of the confidence band
combine_variables: optionally specify a named list where each entry is a character vectors which specifies a group of variables that will be displayed in the same panel. Panel title will be the name of the list element. A function that takes a character vector as an input and produces a list can also be specified (see combine-functions).
ecdf_alpha: the alpha level of the empirical CDF. Can be either a single number or a function taking the number of variables that were combined (when combine_variables is specified) and returns a number. By default, plots showing many ECDFs will have reduced alpha.
...: additional arguments passed to data_for_ecdf_plots(). Most notably, if x is matrix, a max_rank parameter needs to be given.
parameters: DEPRECATED, use variables instead.

Details

arxiv::1903.08008 by A. Vehtari et al.

Arguments

Details

See also