stat_fit_glance: One row summary data frame for a fitted model

Description

stat_fit_glance() fits a model and returns a "tidy" version of the model's fit, using 'glance() methods from packages 'broom', 'broom.mixed', or other sources. The role played by `glance()` is similar to that played by R's anova().

Usage

stat_fit_glance(
  mapping = NULL,
  data = NULL,
  geom = "text_npc",
  position = "identity",
  ...,
  orientation = NA,
  method = "lm",
  method.args = list(formula = y ~ x),
  n.min = 2L,
  fit.seed = NA,
  glance.args = list(),
  label.x = "left",
  label.y = "top",
  hstep = 0,
  vstep = 0.075,
  na.rm = FALSE,
  show.legend = FALSE,
  inherit.aes = TRUE
)

Value

The output of the glance() methods is returned almost as is in the data object, as a data frame. The names of the columns in the returned data are consistent with those returned by method glance()

from package 'broom', that will frequently differ from the name of values returned by the print methods corresponding to the fit or test function used. To explore the values returned by this statistic including the name of variables/columns, which vary depending on the model fitting function and model formula we suggest the use of

geom_debug. An example is shown below.

Arguments

mapping: The aesthetic mapping, usually constructed with aes(). Only needs to be set at the layer level if you are overriding the plot defaults.
data: A layer specific dataset, only needed if you want to override the plot defaults.
geom: The geometric object to use display the data
position: The position adjustment to use for overlapping points on this layer.
...: other arguments passed on to layer. This can include aesthetics whose values you want to set, not map. See layer for more details.
orientation: character Either "x" or "y" controlling the default for formula. The letter indicates the aesthetic considered the explanatory variable in the model fit.
method: function or character If character, "lm", "rlm", "lmrob", "lts", "gls", "ma", "sma", "segreg", "rq" or the name of a model fit function are accepted, possibly followed by the fit function's method argument separated by a colon (e.g. "rlm:M"). If a function is different to lm(), rlm(), ltsReg(), gls(), ma, sma, it must have formal parameters named formula, data, and weights. See Details.
method.args, glance.args: list of named arguments to pass to method and to [generics::glance()], respectively.
n.min: integer Minimum number of distinct values in the explanatory variable (on the rhs of formula) for fitting to the attempted.
fit.seed: RNG seed argument passed to set.seed(). Defaults to NA, indicating that set.seed() should not be called.
label.x, label.y: numeric with range 0..1 "normalized parent coordinates" (npc units) or character if using geom_text_npc() or geom_label_npc(). If using geom_text() or geom_label() numeric in native data units. If too short they will be recycled.
hstep, vstep: numeric in npc units, the horizontal and vertical step used between labels for different groups.
na.rm: a logical indicating whether NA values should be stripped before the computation proceeds.
show.legend: logical. Should this layer be included in the legends? NA, the default, includes if any aesthetics are mapped. FALSE never includes, and TRUE always includes.
inherit.aes: If FALSE, overrides the default aesthetics, rather than combining with them. This is most useful for helper functions that define both data and aesthetics and shouldn't inherit behaviour from the default plot specification, e.g. borders.

Handling of grouping

stat_fit_glance applies the function given by method separately to each group of observations, and factors mapped to aesthetics, including x and y aesthetics, create a separate group for each factor level. Because of this, stat_fit_glance is not useful for annotating plots with results from t.test(), ANOVA or ANCOVA. In such cases use the stat_fit_tb() statistic which applies the model fitting per panel.

Which variables are available for mapping?

Computed variables and their names vary following those used in the specializations of 'broom' methods. A message is issued by default in interactive R sessions, listing the names of all variables accessible by after_stat() within a call to aes() with at least some non-missing values. This default ("colnames") can be changed by setting R option "ggpmisc.stat.vars.message" to "none".

Model formula and model fitting

A ggplot statistic receives as data a data frame that is not the one passed as argument by the user, but instead a data frame with the variables mapped to aesthetics. In stat_poly_eq() the compute function is applied by group, each call "seeing" the subset of data for an individual group. As supported models are for regression lines, variables mapped to x and y should both be continuous, i.e., numeric or date time and model formulas defined using x and y as variable names.

The interpretation of the argument passed to formula is enhanced compared to stat_smooth(). Formulas with x as explanatory variable work as in stat_smooth() but formulas with y as explanatory variable are also accepted. orientation is set automatically based on which explanatory variable appears in the formula. Spline-based smoothers are only partially supported.

Model fit methods supported

Several model fit functions are supported explicitly (see tables), and some of their differences smoothed out. Compatibility is checked late, based on the class of the returned fitted model object. This makes it possible to use wrapper functions that do model selection or other adjustments to the fit procedure on a per panel or per group basis. Moreover, if the value returned as model fit object is NULL or NA, plotting is skipped on a per group within panel basis.

In the case of fitted model objects of classes not explicitly supported, an attempt is made to find the usual accessors and/or fitted object members, and if found, either complete or partial support is frequently achieved. In this case a message is issued encouraging users to check the validity of the values extracted as the structure of fitted model objects belonging to different classes and the values returned by their accessors can vary, potentially resulting in decoding errors leading to the return of wrong values for estimates.

The argument to parameter method can be either the name of a function object, possibly using double colon notation in case its package is not attached, or a character string matching the function name for functions in the search path. This approach makes it possible to support model fit functions that are not dependencies of 'ggpmisc'. Either by attaching the package where the function is defined and passing it by name or as string, or using double colon notation when passing the name of the function.

User-defined functions can be passed as argument to parameter method as long as they have parameters formula, data subset and possibly weights. Additional arguments can be passed to any method as a named list through parameter method.args. As in stat_smooth() prior weights are passed to the model fit functions' weights (plural!) parameter by mapping a numeric variable to plot aesthetic weight (singular!).

Tables 1 lists natively supported model fit functions, with the caveat that only some 'broom' methods' specializations have been actually tested with statistics from 'ggpmisc'. In addition, the statistics based on 'broom' methods require the user to tailor their behaviour by passing additional arguments in the call and occasionally some detective work to find out the names of variables in the returned data frame as these names are set by methods from 'broom'.

Table 1. Model fit methods supported by the different statistics available in package 'ggpmisc'. Column \(f\) indicates whether computations are done by group (G) or by plot panel (P).

Statistic	\(f\)	Supported model fit methods
`stat_poly_line()`	G	"lm", "rlm", "lts", "sma", "ma", "gls", "nls", "onls", others with methods `predict()` or `fitted()`
`stat_poly_eq()`	G	"lm", "rlm", "lts", "sma", "ma", "gls", "nls", "onls", others with needed accesors
`stat_quant_line()`	G	"rq", "rqss"
`stat_quant_band()`	G	"rq", "rqss"
`stat_quant_eq()`	G	"rq", "rqss"
`stat_ma_line()`	G	"SMA", "MA", "RMA", "OLS"
`stat_ma_eq()`	G	"SMA", "MA", "RMA", "OLS"
`stat_fit_residuals()`	G	"lm", "rlm", "lts", "sma", "ma", "gls", "rq", "rqss", "nls", "onls", others with method `residuals()`
`stat_fit_fitted()`	G	"lm", "rlm", "lts", "gls", "rq", "rqss", "nls", "onls", others with method `fitted()`
`stat_fit_deviations()`	G	"lm", "rlm", "lts", "gls", "rq", "rqss", "nls", "onls", others with methods `fitted()` and `weights()`
`stat_fit_augment()`	G	any with 'broom' method `augment()`
`stat_fit_glance()`	G	any with 'broom' method `glance()`
`stat_fit_tidy()`	G	any with 'broom' method `tidy()`
`stat_fit_tb()`	P	any with 'broom' method `tidy()`

The single colon notation is based on parsing the name and is available when passing the name of the fit method as a character string. In a string such as "head:tail" the "head" gives the name of the model fit function and the "tail" gives the argument to pass it's method parameter. This is only a convenience, as method.args can be also used. In some methods, i.e., splines, the default formula = y ~ x needs to be overridden by the user.

Table 2 lists the correspondence of pre-defined method names to model fit method functions. As mentioned above, these are only a subset of the model fit methods that are expected to work. When using these names there is no need for users to attach additional packages but the packages must be available (installed).

Table 2. Available predefined method names, the model fit functions they call, the packages where the functions reside, the class of the returned fitted model object and the arguments that can be passed to their method parameter using single colon notation.

Predefined method names	Model fit methods	R package	Object class
"lm", "lm:qr"	`lm()`	'stats'	"lm"
"rlm", "rlm:M", "rlm:MM"	`rlm()`	'MASS'	"rlm" ("lm")
"lts", "ltsReg"	`ltsReg()`	'robustbase'	"lts"
"ma", "sma", "sma:SMA", "sma:MA", "sma:OLS"	`sma()`	'smatr'	"ma" or "sma"
"gls", "gls:REML", "gls:ML"	`gls()`	'nlme'	"gls"
"rq", "rq:sfn", "rq:sfnc", "rq:lasso"	`rq()`	'quantreg'	"rq"
"rqss", "rqss:sfn", "rqss:sfnc", "rqss:lasso"	`rqss()`	'quantreg'	"rqss"
"SMA", "MA", "RMA", "OLS"	`lmodel2()`	'lmodel2'	("list")

Aesthetics

stat_fit_glance() understands the following aesthetics. Required aesthetics are displayed in bold and defaults are displayed for optional aesthetics:

•	`x`
•	`y`
•	`group`	→ inferred
•	`hjust`	→ `"inward"`
•	`npcx`	→ `after_stat(npcx)`
•	`npcy`	→ `after_stat(npcy)`
•	`vjust`	→ `"inward"`

Learn more about setting these aesthetics in vignette("ggplot2-specs").

Details

stat_fit_glance() together with stat_fit_tidy() and stat_fit_augment(), based on package 'broom' can be used with a broad range of model fitting functions as supported at any given time by package 'broom'. In contrast to stat_poly_eq() which can generate text or expression labels automatically, for these functions the mapping of aesthetic label needs to be explicitly supplied in the call, and labels built on the fly in the mapping to geom aesthetics.

Although arguments passed to parameter glance.args are passed to glance() whether they are silently ignored or obeyed depends on each specialization of glance(), so do carefully read the documentation for the version of glance() corresponding to the method used to fit the model.

This statistics is meant to be used in situations when all of stat_poly_eq(), stat_quant_eq(), stat_ma_eq() and stat_correlation() are unsuitable.

Warning! Not all glance() methods are defined in package 'broom'. glance() specializations for mixed models fits of classes "lme", "nlme", "lme4" and many others are defined in package 'broom.mixed'.

Examples

Run this code

# package 'broom' needs to be installed to run these examples

broom.installed <- requireNamespace("broom", quietly = TRUE)
gginnards.installed <- requireNamespace("gginnards", quietly = TRUE)

if (broom.installed) {
  library(broom)
}

# Regression by panel example
# simpler alternative: stat_poly_eq()
#
if (broom.installed)
  ggplot(mtcars, aes(x = disp, y = mpg)) +
    stat_poly_line() +
    geom_point(aes(colour = factor(cyl))) +
    stat_fit_glance(method = "lm",
                    label.y = "bottom",
                    method.args = list(formula = y ~ x),
                    mapping =
                      aes(label =
                        after_stat(
                          sprintf('italic(r)^2~"="~%.3f~~italic(P)~"="~%.2g',
                                  r.squared, p.value))),
                    parse = TRUE)

# Flipped regression by panel example
# simpler alternative: stat_poly_eq()
#
if (broom.installed)
  ggplot(mtcars, aes(x = disp, y = mpg)) +
    stat_poly_line(orientation = "y") +
    geom_point(aes(colour = factor(cyl))) +
    stat_fit_glance(method = "lm",
                    orientation = "y",
                    label.y = "bottom",
                    method.args = list(formula = y ~ x),
                    mapping =
                      aes(label =
                        after_stat(
                          sprintf('italic(r)^2~"="~%.3f~~italic(P)~"="~%.2g',
                                  r.squared, p.value))),
                    parse = TRUE)

# Regression by group example
# simpler alternative: stat_poly_eq()
#
if (broom.installed)
  ggplot(mtcars, aes(x = disp, y = mpg, colour = factor(cyl))) +
    stat_poly_line() +
    geom_point() +
    stat_fit_glance(method = "lm",
                    label.y = "bottom",
                    method.args = list(formula = y ~ x),
                    mapping = aes(label =
                                after_stat(
                                  sprintf('r^2~"="~%.3f~~italic(P)~"="~%.2g',
                                  r.squared, p.value))),
                    parse = TRUE)

# Weighted regression example
# simpler alternative: stat_poly_eq()
#
if (broom.installed)
  ggplot(mtcars, aes(x = disp, y = mpg, weight = cyl)) +
    stat_poly_line() +
    geom_point(aes(colour = factor(cyl))) +
    stat_fit_glance(method = "lm",
                    label.y = "bottom",
                    method.args = list(formula = y ~ x,
                                       weights = quote(weight)),
                    mapping = aes(label =
                      after_stat(
                        sprintf('r^2~"="~%.3f~~italic(P)~"="~%.2g',
                                r.squared, p.value))),
                    parse = TRUE)

# correlation test
# simpler alternative: stat_correlation()
#
if (broom.installed)
  ggplot(mtcars, aes(x = disp, y = mpg)) +
    geom_point() +
    stat_fit_glance(method = "cor.test",
                    label.y = "bottom",
                    method.args = list(formula = ~ x + y),
                    mapping = aes(label =
                      after_stat(
                        sprintf('r[Pearson]~"="~%.3f~~italic(P)~"="~%.2g',
                                 estimate, p.value))),
                    parse = TRUE)

if (broom.installed)
  ggplot(mtcars, aes(x = disp, y = mpg)) +
    geom_point() +
    stat_fit_glance(method = "cor.test",
                    label.y = "bottom",
                    method.args = list(formula = ~ x + y,
                                       method = "spearman",
                                       exact = FALSE),
                    mapping = aes(label =
                      after_stat(
                        sprintf('r[Spearman]~"="~%.3f~~italic(P)~"="~%.2g',
                                estimate, p.value))),
                    parse = TRUE)

# Inspecting the returned data using geom_debug_group()
#
if (gginnards.installed) {
    library(gginnards)
}

if (broom.installed && gginnards.installed) {
    ggplot(mtcars, aes(x = disp, y = mpg)) +
      stat_smooth(method = "lm") +
      geom_point(aes(colour = factor(cyl))) +
      stat_fit_glance(method = "lm",
                      method.args = list(formula = y ~ x),
                      geom = "debug_group")
}

Run the code above in your browser using DataLab