Skip to contents

Augment Model Data for Diagnostic Plots

Source: R/lm_plot.df.R
lm_plot.df.Rd

Generates an augmented data frame from a linear model object, including fitted values, residuals, leverage, Cook's distance, prediction intervals, and outlier/influence flags. This function prepares model diagnostics for plotting.

Usage

lm_plot.df(mdl)

Arguments

mdl

An object of class lm, representing the fitted linear model.

Value

A data frame with augmented diagnostic variables, one row per observation.

Details

The returned data frame contains key statistics for each observation:

  • .id: Observation identifier

  • .sequence: Sequence index

  • .fits: Fitted/predicted values

  • .resid: Residuals

  • .obs: Observed values

  • .sigma: Estimated standard deviation for each residual

  • .std.resid: Standardized residuals

  • .stud.resid: Studentized residuals

  • .lower.pi, .upper.pi: Lower/upper 95% prediction interval bounds

  • .cooksd: Cook's distance

  • .hat: Leverage (diagonal of the hat matrix)

  • .cooksd.is.infl, .hat.is.infl: Logical flags for influential points

  • .quantile: Theoretical normal quantile of residuals

  • outlier: Flag for ordinary residual outlier ("outl" or "reg")

  • .stud.outl: Flag for studentized residual influential point ("infl" or "reg")

See also

influence, influence.measures, rstandard, rstudent, outlier

Examples

mdl <- lm(Sepal.Length ~ Sepal.Width, data = iris)
df <- lm_plot.df(mdl)
head(df)
#>   .id .sequence    .fits     .resid .obs    .sigma .std.resid .stud.resid
#> 1   1         1 5.744459 -0.6444588  5.1 0.8261667 -0.7864325  -0.7854139
#> 2   2         2 5.856139 -0.9561394  4.9 0.8241081 -1.1627712  -1.1641660
#> 3   3         3 5.811467 -1.1114672  4.7 0.8227693 -1.3520774  -1.3559020
#> 4   4         4 5.833803 -1.2338033  4.6 0.8215777 -1.5004021  -1.5068286
#> 5   5         5 5.722123 -0.7221227  5.0 0.8257158 -0.8827646  -0.8821026
#> 6   6         6 5.655114 -0.2551144  5.4 0.8276221 -0.3142222  -0.3132634
#>   .lower.pi .upper.pi     .cooksd        .hat .cooksd.is.infl .hat.is.infl
#> 1  4.102926  7.385992 0.004260169 0.013589132           FALSE        FALSE
#> 2  4.220127  7.492152 0.004616605 0.006782791           FALSE        FALSE
#> 3  4.174965  7.447969 0.006801185 0.007385705           FALSE        FALSE
#> 4  4.197833  7.469773 0.007627752 0.006730978           FALSE        FALSE
#> 5  4.077773  7.366472 0.006766613 0.017070023           FALSE        FALSE
#> 6  3.998939  7.311290 0.001618927 0.031751938           FALSE        FALSE
#>    .quantile outlier .stud.outl
#> 1 -0.6744898     reg        reg
#> 2 -1.0954185     reg        reg
#> 3 -1.5273795     reg        reg
#> 4 -1.7907506     reg        reg
#> 5 -0.7609839     reg        reg
#> 6 -0.2104284     reg        reg