descsuppRplots: Generate Plots for All Variables in Descriptive Tables

Description

Visualizes variables from descriptive tables produced by 'descsuppR::buildDescrTbl()' using 'ggstatsplot'. It automatically maps each variable to a suitable 'ggstatsplot' plotting function based on the applied or suggested statistical test. Users can override the automatic mapping via a named list of plot specifications. The package supports grouped and ungrouped tables, and forwards additional arguments to the underlying 'ggstatsplot' functions, providing quick, reproducible, and customizable default visualizations for descriptive summaries.

Details

Take the output of buildDescrTbl from the descsuppR package and use ggstatsplot to generate some default visualizations.

Author(s)

Maintainer: Fabian Kück fabian.kueck@med.uni-goettingen.de

Authors:

• Andreas Leha andreas.leha@med.uni-goettingen.de

Report default plot functions

Description

Report default plot functions

Usage

get_plotfun_defaults()

Value

tibble with the currently set default suggestions for plotting function

Author(s)

Dr. Andreas Leha

Examples

get_plotfun_defaults()

Variable Visualizations for Descriptive Tables

Description

Using as underlying functions the plot functions from ggstatsplot this package visualizes the variables in a descriptive table generated via buildDescrTbl.

Usage

plotDescTbl(
  dtab,
  plotfuns = NULL,
  verbose = 0,
  pairwise.display = "none",
  pval.display = FALSE,
  ...
)

Arguments

Value

tibble with column p containing the plots.

Author(s)

Dr. Andreas Leha

Examples

ttt <- data.frame(data="training set",
                  age=runif(100, 0, 100),
                  sex=as.factor(sample(c("m","f"), 100, replace=TRUE, prob=c(0.3, 0.7))),
                  score=factor(sample(1:5, 100, replace=TRUE),
                    ordered=TRUE,
                    levels=1:5))
ttt2 <- data.frame(data="test set",
                   age=runif(100, 0, 100),
                   sex=as.factor(sample(c("m","f"), 100, replace=TRUE, prob=c(0.5,0.5))),
                   score=factor(sample(1:5, 100, replace=TRUE),
                     ordered=TRUE,
                     levels=1:5))

## ungrouped table and plots
dtab <- buildDescrTbl(dplyr::mutate(rbind(ttt, ttt2), data = as.factor(data)),
                      includeNAs=TRUE,
                      dopvals = TRUE)

plotDescTbl(dtab)

## grouped table and plots
dtab <- buildDescrTbl(dplyr::mutate(rbind(ttt, ttt2), data = as.factor(data)),
                      groupby="data",
                      includeNAs=TRUE,
                      dopvals = TRUE)

plotDescTbl(dtab)

Change default plot functions

Description

Change default plot functions

Usage

set_plotfun_defaults(df)

Arguments

Value

invisible. tibble with the old (prior to the change) set default suggestions for plotting function

Author(s)

Dr. Andreas Leha

Examples

current <- get_plotfun_defaults()
current$plotfun[current$test == "Fisher's Exact Test for Count Data"] <- "emptyplot"
current$args[current$test == "Fisher's Exact Test for Count Data"] <- list(NULL)
set_plotfun_defaults(current)

Function suggesting a default plot for a given parameter in a descriptive table

Description

The plot suggestion is based on the test that was used for that variable.

Usage

suggestPlot(ltest)

Arguments

Details

The default choice is taken from plotDescTbl_options$plotfun_defaults with the current associations.

If no test was performed than numeric variables are treated as if tested via t-tests and non-numeric variables as if tested via Fisher test.

Value

tibble with columns plotfun an args where plotfun is a string with the name of the function to be called to generate the plot and args is a list of additional (besides the data) arguments

Author(s)

Dr. Andreas Leha

Examples

suggestPlot("Fisher's Exact Test for Count Data")

suggestPlot("foobar")

Wrapper for ggplot2::ggplot()

Description

Wrapper for ggplot2::ggplot()

Usage

w.emptyplot(data, parameter, groupby = NULL, pval.display = FALSE, ldat, ...)

Arguments

Value

A ggplot object.

Author(s)

Dr. Andreas Leha

Examples

w.emptyplot()

Wrapper for ggstatsplot::ggbarstats with optional p-value annotation

Description

This function creates a 'ggbarstats' plot from 'ggstatsplot', and optionally adds a p-value annotation using 'geom_signif' for two-group comparisons based on values in 'ldat'.

Usage

w.ggbarstats(
  data,
  parameter,
  groupby = NULL,
  pval.display = FALSE,
  ldat,
  tip_length = 4e-04,
  pval_y_position = 1.05,
  ...
)

Arguments

Details

If groupby is not NULL and there are exactly two levels in the grouping variable, and pval.display = TRUE, a p-value from ldat is added between the two groups using geom_signif. If "adjusted p value" is present in ldat, it is used; otherwise, "p value" is used.

Value

A ggplot object.

Author(s)

Fabian Kück

Examples

## ungrouped plot
p <- w.ggbarstats(mtcars, "gear",
                  bf.message = FALSE,
                  proportion.test = FALSE)
if (FALSE) print(p)

## grouped plot
p <- w.ggbarstats(mtcars, "gear", groupby = "cyl",
                  bf.message = FALSE,
                  proportion.test = FALSE)
if (FALSE) print(p)

## grouped plot with p value annotation for >2 groups
pval <- fisher.test(mtcars$gear, mtcars$cyl)$p.value
pval <- descutils::prettyPvalues(pval, digits = 3, orgbold = FALSE, lhs = "p")
ldat <- data.frame(`p value` = pval, check.names = FALSE)
p <- w.ggbarstats(mtcars, "gear", groupby = "cyl", pval.display = TRUE, ldat = ldat,
                  bf.message = FALSE,
                  proportion.test = FALSE)
if (FALSE) print(p)

## grouped plot with p value annotation for two groups
pval <- fisher.test(mtcars$gear, mtcars$vs)$p.value
pval <- descutils::prettyPvalues(pval, digits = 3, orgbold = FALSE, lhs = "p")
ldat <- data.frame(`p value` = pval, check.names = FALSE)
p <- w.ggbarstats(mtcars, "gear", groupby = "vs", pval.display = TRUE, ldat = ldat,
                  bf.message = FALSE,
                  proportion.test = FALSE)
if (FALSE) print(p)

Wrapper for ggstatsplot::ggbetweenstats with optional p-value annotation and date formatting

Description

This function creates a 'ggbetweenstats' plot from 'ggstatsplot', and optionally adds a p-value annotation using 'geom_signif' for two-group comparisons based on values in 'ldat'. If requested, it can also format the y-axis as dates.

Usage

w.ggbetweenstats(
  data,
  parameter,
  groupby = NULL,
  pval.display = FALSE,
  ldat,
  ...
)

Arguments

Details

If groupby is not NULL and there are exactly two levels in the grouping variable, and pval.display = TRUE, a p-value from ldat is added between the two groups using geom_signif. If "adjusted p value" is present in ldat, it is used; otherwise, "p value" is used. If wasDate is TRUE, the y-axis labels are formatted as dates.

Value

A ggplot object.

Author(s)

Fabian Kück

Examples

## ungrouped plot
p <- w.ggbetweenstats(iris, "Sepal.Length")
if (FALSE) print(p)

## grouped plot
p <- w.ggbetweenstats(iris, "Sepal.Length", groupby = "Species",
                      bf.message = FALSE, pairwise.display = "none")
if (FALSE) print(p)

## grouped plot with p value annotation for >2 groups
pval <- anova(lm(Sepal.Length ~ Species, data = iris))['Species', 'Pr(>F)']
pval <- descutils::prettyPvalues(pval, digits = 3, orgbold = FALSE, lhs = "p")
ldat <- data.frame(`p value` = pval, check.names = FALSE)
p <- w.ggbetweenstats(iris, "Sepal.Length", groupby = "Species", pval.display = TRUE, ldat = ldat,
                      bf.message = FALSE, pairwise.display = "none")
if (FALSE) print(p)

## setup simple data with only two groups
iris2 <- iris[iris$Species %in% c("setosa", "virginica"),]

## grouped plot with p value annotation for two groups
pval <- t.test(Sepal.Length ~ Species, data = iris2)$p.value
pval <- descutils::prettyPvalues(pval, digits = 3, orgbold = FALSE, lhs = "p")
ldat <- data.frame(`p value` = pval, check.names = FALSE)
p <- w.ggbetweenstats(iris2, "Sepal.Length", groupby = "Species", pval.display = TRUE, ldat = ldat,
                      bf.message = FALSE, pairwise.display = "none")
if (FALSE) print(p)

Wrapper for ggstatsplot::ggscatterstats with optional p-value annotation

Description

This function creates a 'ggscatterstats' plot from 'ggstatsplot', and optionally adds a p-value annotation based on values in 'ldat'. It can also format the y-axis as dates.

Usage

w.ggscatterstats(data, parameter, groupby, pval.display = FALSE, ldat, ...)

Arguments

Details

If groupby is not NULL and there are exactly two levels in the grouping variable, and pval.display = TRUE, a p-value from ldat is added between the two groups using geom_signif. If "adjusted p value" is present in ldat, it is used; otherwise, "p value" is used. If wasDate is TRUE, the y-axis labels are formatted as dates.

Value

A ggplot object.

Author(s)

Fabian Kück

Examples

## plot
p <- w.ggscatterstats(mtcars, parameter = "cyl", groupby = "carb",
                      bf.message = FALSE,)
if (FALSE) print(p)

## plot with p value annotation
pval <- cor.test(~ cyl + carb, data = mtcars, method = "kendall")$p.value
pval <- descutils::prettyPvalues(pval, digits = 3, orgbold = FALSE, lhs = "p")
ldat <- data.frame(`p value` = pval, check.names = FALSE)
p <- w.ggscatterstats(mtcars, parameter = "cyl", groupby = "carb", pval.display = TRUE, ldat = ldat,
                      bf.message = FALSE,)
if (FALSE) print(p)