PPC intervals

Medians and central interval estimates of yrep with y overlaid. See the Plot Descriptions section, below.

ppc_intervals(
  y,
  yrep,
  x = NULL,
  ...,
  prob = 0.5,
  prob_outer = 0.9,
  size = 1,
  fatten = 3
)

ppc_intervals_grouped(
  y,
  yrep,
  x = NULL,
  group,
  ...,
  facet_args = list(),
  prob = 0.5,
  prob_outer = 0.9,
  size = 1,
  fatten = 3
)

ppc_ribbon(
  y,
  yrep,
  x = NULL,
  ...,
  prob = 0.5,
  prob_outer = 0.9,
  alpha = 0.33,
  size = 0.25
)

ppc_ribbon_grouped(
  y,
  yrep,
  x = NULL,
  group,
  ...,
  facet_args = list(),
  prob = 0.5,
  prob_outer = 0.9,
  alpha = 0.33,
  size = 0.25
)

ppc_intervals_data(
  y,
  yrep,
  x = NULL,
  group = NULL,
  ...,
  prob = 0.5,
  prob_outer = 0.9
)

ppc_ribbon_data(
  y,
  yrep,
  x = NULL,
  group = NULL,
  ...,
  prob = 0.5,
  prob_outer = 0.9
)

Arguments

y	A vector of observations. See Details.
yrep	An \(S\) by \(N\) matrix of draws from the posterior predictive distribution, where \(S\) is the size of the posterior sample (or subset of the posterior sample used to generate `yrep`) and \(N\) is the number of observations (the length of `y`). The columns of `yrep` should be in the same order as the data points in `y` for the plots to make sense. See Details for additional instructions.
x	A numeric vector the same length as `y` to use as the x-axis variable. For example, `x` could be a predictor variable from a regression model, a time variable for time-series models, etc. If `x` is missing or `NULL`, then `1:length(y)` is used for the x-axis.
...	Currently unused.
prob, prob_outer	Values between 0 and 1 indicating the desired probability mass to include in the inner and outer intervals. The defaults are `prob=0.5` and `prob_outer=0.9`.
group	A grouping variable (a vector or factor) the same length as `y`. Each value in `group` is interpreted as the group level pertaining to the corresponding value of `y`.
facet_args	An optional list of arguments (other than `facets`) passed to `ggplot2::facet_wrap()` to control faceting.
alpha, size, fatten	Arguments passed to geoms. For ribbon plots `alpha` and `size` are passed to `ggplot2::geom_ribbon()`. For interval plots `size` and `fatten` are passed to `ggplot2::geom_pointrange()`.

Value

The plotting functions return a ggplot object that can be further customized using the ggplot2 package. The functions with suffix _data() return the data that would have been drawn by the plotting function.

Plot Descriptions

ppc_intervals(), ppc_ribbon()

100*prob\ value. ppc_intervals() plots intervals as vertical bars with points indicating yrep medians and darker points indicating observed y values. ppc_ribbon() plots a ribbon of connected intervals with a line through the median of yrep and a darker line connecting observed y values. In both cases an optional x variable can also be specified for the x-axis variable.Depending on the number of observations and the variability in the predictions at different values of x, one or the other of these plots may be easier to read than the other.

ppc_intervals_grouped(), ppc_ribbon_grouped()

Same as ppc_intervals() and ppc_ribbon(), respectively, but a separate plot (facet) is generated for each level of a grouping variable.

References

Gelman, A., Carlin, J. B., Stern, H. S., Dunson, D. B., Vehtari, A., and Rubin, D. B. (2013). Bayesian Data Analysis. Chapman & Hall/CRC Press, London, third edition. (Ch. 6)

Examples

y <- rnorm(50)
yrep <- matrix(rnorm(5000, 0, 2), ncol = 50)

color_scheme_set("brightblue")
ppc_ribbon(y, yrep)
ppc_intervals(y, yrep)

# change x axis to y values (instead of indices) and add x = y line
ppc_intervals(y, yrep, x = y) + abline_01()


color_scheme_set("teal")
year <- 1950:1999
ppc_ribbon(y, yrep, x = year, alpha = 0, size = 0.75) + ggplot2::xlab("Year")

color_scheme_set("pink")
year <- rep(2000:2009, each = 5)
group <- gl(5, 1, length = 50, labels = LETTERS[1:5])
ppc_ribbon_grouped(y, yrep, x = year, group) +
  ggplot2::scale_x_continuous(breaks = pretty)

ppc_ribbon_grouped(
 y, yrep, x = year, group,
 facet_args = list(scales = "fixed"),
 alpha = 1,
 size = 2
) +
 xaxis_text(FALSE) +
 xaxis_ticks(FALSE) +
 panel_bg(fill = "gray20")

ppc_dat <- ppc_intervals_data(y, yrep, x = year, prob = 0.5)
ppc_group_dat <- ppc_intervals_data(y, yrep, x = year, group = group, prob = 0.5)

# \dontrun{
library("rstanarm")
fit <- stan_glmer(mpg ~ wt + (1|cyl), data = mtcars, refresh = 0)
#> Warning: There were 1 divergent transitions after warmup. Increasing adapt_delta above 0.95 may help. See
#> http://mc-stan.org/misc/warnings.html#divergent-transitions-after-warmup
#> Warning: Examine the pairs() plot to diagnose sampling problems
yrep <- posterior_predict(fit)

color_scheme_set("purple")
with(mtcars, ppc_intervals(mpg, yrep, x = wt, prob = 0.5)) +
 panel_bg(fill="gray90", color = NA) +
 grid_lines(color = "white")

ppc_intervals_grouped(y = mtcars$mpg, yrep, prob = 0.8,
                      x = mtcars$wt, group = mtcars$cyl)


color_scheme_set("gray")
ppc_intervals(mtcars$mpg, yrep, prob = 0.5) +
 ggplot2::scale_x_continuous(
   labels = rownames(mtcars),
   breaks = 1:nrow(mtcars)
 ) +
 xaxis_text(angle = -70, vjust = 1, hjust = 0)

# }

Arguments

Value

Plot Descriptions

References

See also

Examples