1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
|
% Generated by roxygen2: do not edit by hand
% Please edit documentation in R/ppd-test-statistics.R
\name{PPD-test-statistics}
\alias{PPD-test-statistics}
\alias{PPD-statistics}
\alias{ppd_stat}
\alias{ppd_stat_grouped}
\alias{ppd_stat_freqpoly}
\alias{ppd_stat_freqpoly_grouped}
\alias{ppd_stat_2d}
\alias{ppd_stat_data}
\title{PPD test statistics}
\usage{
ppd_stat(
ypred,
stat = "mean",
...,
binwidth = NULL,
bins = NULL,
breaks = NULL,
freq = TRUE
)
ppd_stat_grouped(
ypred,
group,
stat = "mean",
...,
facet_args = list(),
binwidth = NULL,
bins = NULL,
breaks = NULL,
freq = TRUE
)
ppd_stat_freqpoly(
ypred,
stat = "mean",
...,
facet_args = list(),
binwidth = NULL,
bins = NULL,
freq = TRUE
)
ppd_stat_freqpoly_grouped(
ypred,
group,
stat = "mean",
...,
facet_args = list(),
binwidth = NULL,
bins = NULL,
freq = TRUE
)
ppd_stat_2d(ypred, stat = c("mean", "sd"), ..., size = 2.5, alpha = 0.7)
ppd_stat_data(ypred, group = NULL, stat)
}
\arguments{
\item{ypred}{An \code{S} by \code{N} matrix of draws from the posterior (or prior)
predictive distribution. The number of rows, \code{S}, is the size of the
posterior (or prior) sample used to generate \code{ypred}. The number of
columns, \code{N}, is the number of predicted observations.}
\item{stat}{A single function or a string naming a function, except for the
2D plot which requires a vector of exactly two names or functions. In all
cases the function(s) should take a vector input and return a scalar
statistic. If specified as a string (or strings) then the legend will
display the function name(s). If specified as a function (or functions)
then generic naming is used in the legend.}
\item{...}{Currently unused.}
\item{binwidth}{Passed to \code{\link[ggplot2:geom_histogram]{ggplot2::geom_histogram()}} to override
the default binwidth.}
\item{bins}{Passed to \code{\link[ggplot2:geom_histogram]{ggplot2::geom_histogram()}} to override
the default binwidth.}
\item{breaks}{Passed to \code{\link[ggplot2:geom_histogram]{ggplot2::geom_histogram()}} as an
alternative to \code{binwidth}.}
\item{freq}{For histograms, \code{freq=TRUE} (the default) puts count on the
y-axis. Setting \code{freq=FALSE} puts density on the y-axis. (For many
plots the y-axis text is off by default. To view the count or density
labels on the y-axis see the \code{\link[=yaxis_text]{yaxis_text()}} convenience
function.)}
\item{group}{A grouping variable of the same length as \code{y}.
Will be coerced to \link[base:factor]{factor} if not already a factor.
Each value in \code{group} is interpreted as the group level pertaining
to the corresponding observation.}
\item{facet_args}{A named list of arguments (other than \code{facets}) passed
to \code{\link[ggplot2:facet_wrap]{ggplot2::facet_wrap()}} or \code{\link[ggplot2:facet_grid]{ggplot2::facet_grid()}}
to control faceting. Note: if \code{scales} is not included in \code{facet_args}
then \strong{bayesplot} may use \code{scales="free"} as the default (depending
on the plot) instead of the \strong{ggplot2} default of \code{scales="fixed"}.}
\item{size, alpha}{For the 2D plot only, arguments passed to
\code{\link[ggplot2:geom_point]{ggplot2::geom_point()}} to control the appearance of scatterplot points.}
}
\value{
The plotting functions return a ggplot object that can be further
customized using the \strong{ggplot2} package. The functions with suffix
\verb{_data()} return the data that would have been drawn by the plotting
function.
}
\description{
The distribution of a (test) statistic \code{T(ypred)}, or a pair of (test)
statistics, over the simulations from the posterior or prior predictive
distribution. Each of these functions makes the same plot as the
corresponding \code{\link[=PPC-test-statistics]{ppc_}} function but without comparing to
any observed data \code{y}. The \strong{Plot Descriptions} section at
\link{PPC-test-statistics} has details on the individual plots.
}
\details{
For Binomial data, the plots may be more useful if
the input contains the "success" \emph{proportions} (not discrete
"success" or "failure" counts).
}
\examples{
yrep <- example_yrep_draws()
ppd_stat(yrep)
ppd_stat(yrep, stat = "sd") + legend_none()
# use your own function for the 'stat' argument
color_scheme_set("brightblue")
q25 <- function(y) quantile(y, 0.25)
ppd_stat(yrep, stat = "q25") # legend includes function name
}
\references{
Gabry, J. , Simpson, D. , Vehtari, A. , Betancourt, M. and
Gelman, A. (2019), Visualization in Bayesian workflow.
\emph{J. R. Stat. Soc. A}, 182: 389-402. doi:10.1111/rssa.12378.
(\href{https://rss.onlinelibrary.wiley.com/doi/full/10.1111/rssa.12378}{journal version},
\href{https://arxiv.org/abs/1709.01449}{arXiv preprint},
\href{https://github.com/jgabry/bayes-vis-paper}{code on GitHub})
}
\seealso{
Other PPDs:
\code{\link{PPD-distributions}},
\code{\link{PPD-intervals}},
\code{\link{PPD-overview}}
}
\concept{PPDs}
|
