Visualizations with statistical details: The 'ggstatsplot' approach

Indrajeet Patil

doi:10.21105/joss.03167

Scatterplots from {ggplot2} combined with marginal distribution plots with statistical details.

Usage

ggscatterstats(
  data,
  x,
  y,
  type = "parametric",
  conf.level = 0.95,
  bf.prior = 0.707,
  bf.message = TRUE,
  tr = 0.2,
  digits = 2L,
  results.subtitle = TRUE,
  label.var = NULL,
  label.expression = NULL,
  marginal = TRUE,
  point.args = list(size = 3, alpha = 0.4, stroke = 0),
  point.width.jitter = 0,
  point.height.jitter = 0,
  point.label.args = list(size = 3, max.overlaps = 1e+06),
  smooth.line.args = list(linewidth = 1.5, color = "blue", method = "lm", formula = y ~
    x),
  xsidehistogram.args = list(fill = "#4285F4", color = "black", na.rm = TRUE),
  ysidehistogram.args = list(fill = "#EA4335", color = "black", na.rm = TRUE),
  xsidehistogram.scale = list(),
  ysidehistogram.scale = list(),
  xlab = NULL,
  ylab = NULL,
  title = NULL,
  subtitle = NULL,
  caption = NULL,
  ggtheme = ggstatsplot::theme_ggstatsplot(),
  ggplot.component = NULL,
  ...
)

Arguments

data

A data frame (or a tibble) from which variables specified are to be taken. Other data types (e.g., matrix, table, array, etc.) will not be accepted. Additionally, grouped data frames from {dplyr} should be ungrouped before they are entered as data.

x

The column in data containing the explanatory variable to be plotted on the x-axis.

y

The column in data containing the response (outcome) variable to be plotted on the y-axis.

type

A character specifying the type of statistical approach:

"parametric"
"nonparametric"
"robust"
"bayes"

You can specify just the initial letter (e.g. "np" or "bf"). Matching is on the initial lowercase letter only, so any other value (including capitalized values such as "Bayes") falls back to "parametric" without a warning.

conf.level

Scalar between 0 and 1 (default: 95% confidence/credible intervals, 0.95).

bf.prior

A number between 0.5 and 2 (default 0.707), the prior width to use in calculating Bayes factors and posterior estimates. In addition to numeric arguments, several named values are also recognized: "medium", "wide", and "ultrawide", corresponding to r scale values of 1/2, sqrt(2)/2, and 1, respectively. In case of an ANOVA, this value corresponds to scale for fixed effects.

bf.message

Logical that decides whether to display the Bayes Factor in favor of the null hypothesis as a caption. Relevant only for parametric tests (Default: TRUE).

tr

Trim level for the mean when carrying out robust tests. In case of an error, try reducing the value of tr, which is by default set to 0.2. Lowering the value might help.

digits

Number of digits for rounding or significant figures. May also be "signif" to return significant figures or "scientific" to return scientific notation. Control the number of digits by adding the value as suffix, e.g. digits = "scientific4" to have scientific notation with 4 decimal places, or digits = "signif5" for 5 significant figures (see also signif()).

results.subtitle

Logical that decides whether the results of statistical tests are displayed as a subtitle (Default: TRUE). Setting this to FALSE also skips the automatic Bayes Factor caption. Other plot details have their own arguments.

label.var

Variable to use for point labels entered as a symbol (e.g. var1).

label.expression

An expression evaluating to a logical vector that determines the subset of data points to label (e.g. y < 4 & z < 20). While using this argument with purrr::pmap(), you will have to provide a quoted expression (e.g. quote(y < 4 & z < 20)).

marginal

Logical that decides whether marginal distributions are plotted on the axes using {ggside} (Default: TRUE).

point.args

A list of additional aesthetic arguments passed to ggplot2::geom_point().

point.width.jitter, point.height.jitter

Amount of jitter (in data units) in x and y direction, respectively, passed to ggplot2::position_jitter() (Default: 0, i.e., no jitter). Do not specify the jitter in point.args, because it is shared by the points and their labels.

point.label.args

A list of additional aesthetic arguments passed to ggrepel::geom_label_repel(), which displays the point labels.

smooth.line.args

A list of additional aesthetic arguments passed to ggplot2::geom_smooth(), which displays the regression line. Its confidence band uses conf.level.

xsidehistogram.args, ysidehistogram.args

A list of arguments passed to ggside::geom_xsidehistogram() and ggside::geom_ysidehistogram(), respectively, to change the marginal histograms.

xsidehistogram.scale, ysidehistogram.scale

A list of arguments passed to ggside::scale_xsidey_continuous() and ggside::scale_ysidex_continuous(), respectively, to control the scale of marginal histograms (e.g., breaks, limits, transform). Default is list() (no modifications).

xlab

Label for x axis variable. If NULL (default), variable name for x will be used.

ylab

Label for y axis variable. If NULL (default), variable name for y will be used.

title

The text for the plot title.

subtitle

The text for the plot subtitle. Used only if results.subtitle = FALSE.

caption

The text for the plot caption. It will be replaced by the Bayes Factor caption whenever that is displayed (see bf.message).

ggtheme

A {ggplot2} theme (Default: theme_ggstatsplot()). Any {ggplot2} theme (e.g., ggplot2::theme_bw()) or theme from an extension package (e.g., ggthemes::theme_fivethirtyeight()) can be used. Note that some themes remove plot details. For example, ggthemes::theme_fivethirtyeight() removes the secondary Y-axis, which ggbetweenstats() uses to show details about pairwise comparisons.

ggplot.component

A {ggplot2} component, or a list of components, to add to the plot (Default: NULL). This is mostly helpful for the grouped_ variants, where it is applied to each individual plot.

...

Additional arguments (currently ignored).

Value

A ggplot object, which can be further modified with {ggplot2} functions. Use extract_stats() to get the statistical details shown in the plot as data frames.

Details

Rows with a missing value in either x or y are removed before plotting and analysis.

For details, see: https://www.indrapatil.com/ggstatsplot/articles/web_only/ggscatterstats.html

Note

Point labels use ggrepel::geom_label_repel() to keep them from overlapping, so plotting can become very slow (and the plot file large) if many labels overlap.

Summary of graphics

graphical element	`geom` used	argument for further modification
raw data	`ggplot2::geom_point()`	`point.args`
labels for raw data	`ggrepel::geom_label_repel()`	`point.label.args`
smooth line	`ggplot2::geom_smooth()`	`smooth.line.args`
marginal histograms	`ggside::geom_xsidehistogram()`, `ggside::geom_ysidehistogram()`	`xsidehistogram.args`, `ysidehistogram.args`

Correlation analyses

Below is a summary of:

the statistical test carried out for inferential statistics
the type of effect size estimate and a measure of uncertainty for this estimate
the functions used internally to compute these details

Hypothesis testing and effect size estimation

Type	Test	CI available?	Function used
Parametric	Pearson's correlation coefficient	Yes	`correlation::correlation()`
Non-parametric	Spearman's rank correlation coefficient	Yes	`correlation::correlation()`
Robust	Winsorized Pearson's correlation coefficient	Yes	`correlation::correlation()`
Bayesian	Bayesian Pearson's correlation coefficient	Yes	`correlation::correlation()`

Examples

set.seed(123)

# creating a plot
p <- ggscatterstats(
  iris,
  x = Sepal.Width,
  y = Petal.Length,
  label.var = Species,
  label.expression = Sepal.Length > 7.6
) +
  ggplot2::geom_rug(sides = "b")

# looking at the plot
p


# extracting details from statistical tests
extract_stats(p)
#> $subtitle_data
#> # A tibble: 1 × 14
#>   parameter1  parameter2   effectsize          estimate conf.level conf.low
#>   <chr>       <chr>        <chr>                  <dbl>      <dbl>    <dbl>
#> 1 Sepal.Width Petal.Length Pearson correlation   -0.428       0.95   -0.551
#>   conf.high statistic df.error      p.value method              n.obs
#>       <dbl>     <dbl>    <int>        <dbl> <chr>               <int>
#> 1    -0.288     -5.77      148 0.0000000451 Pearson correlation   150
#>   conf.method expression
#>   <chr>       <list>    
#> 1 normal      <language>
#> 
#> $caption_data
#> # A tibble: 1 × 17
#>   parameter1  parameter2   effectsize                   estimate conf.level
#>   <chr>       <chr>        <chr>                           <dbl>      <dbl>
#> 1 Sepal.Width Petal.Length Bayesian Pearson correlation   -0.422       0.95
#>   conf.low conf.high    pd rope.percentage prior.distribution prior.location
#>      <dbl>     <dbl> <dbl>           <dbl> <chr>                       <dbl>
#> 1   -0.551    -0.290     1               0 beta                         1.41
#>   prior.scale    bf10 method                       n.obs conf.method expression
#>         <dbl>   <dbl> <chr>                        <int> <chr>       <list>    
#> 1        1.41 312665. Bayesian Pearson correlation   150 HDI         <language>
#> 
#> $pairwise_comparisons_data
#> NULL
#> 
#> $descriptive_data
#> NULL
#> 
#> $one_sample_data
#> NULL
#> 
#> $tidy_data
#> NULL
#> 
#> $glance_data
#> NULL
#> 
#> attr(,"class")
#> [1] "ggstatsplot_stats" "list"             

# customize marginal histogram bins and scales
ggscatterstats(
  mtcars,
  x = wt,
  y = mpg,
  results.subtitle = FALSE,
  xsidehistogram.args = list(fill = "#4285F4", color = "black", na.rm = TRUE, binwidth = 0.5),
  ysidehistogram.args = list(fill = "#EA4335", color = "black", na.rm = TRUE, bins = 15),
  xsidehistogram.scale = list(breaks = seq(0, 15, 5)),
  ysidehistogram.scale = list(breaks = seq(0, 15, 5))
)

Scatterplot with marginal distributions and statistical results