Skip to contents

Plot Residuals vs. Observation Order (Autocorrelation Check)

Source: R/lm_plot.ac.R
lm_plot.ac.Rd

Creates a plot of residuals against the sequence/order of observations to visually assess independence and detect autocorrelation. Optionally overlays results from the Durbin–Watson test and labels outliers.

Usage

lm_plot.ac(
  mdl,
  opt = list(),
  parm = list(),
  df = lm_plot.df(mdl),
  plts = list()
)

Arguments

mdl

A fitted model object (typically from lm).

opt

List of options, where pval.DW (logical, default = FALSE) indicates whether to include Durbin-Watson p-value on the plot.

parm

A list of plotting parameters, usually from lm_plot.parms().

df

Data frame with augmented model data. Defaults to lm_plot.df(mdl).

plts

A list of ggplot objects to which this plot will be added.

Value

A list containing:

  • mdl Fitted model object,

  • opt Options used, including pval.DW,

  • parm Parameter list with Durbin-Watson test results added,

  • df Data frame used for plotting,

  • plts List of ggplot objects, including the $ac element.

Details

Points are colored and shaped according to whether they are residual outliers (as determined by Tukey's boxplot rule). The function can label points using ggrepel if parm$pts$id$outl or parm$pts$id$reg are set to TRUE.

Examples

if (FALSE) { # \dontrun{
fit <- lm(mpg ~ wt + hp, data = mtcars)
lm_plot.ac(fit)
} # }