Skip to contents

Residual vs fitted plot

Source: R/ols-residual-vs-predicted-plot.R
ols_plot_resid_fit.Rd

Scatter plot of residuals on the y axis and fitted values on the x axis to detect non-linearity, unequal error variances, and outliers.

Usage

ols_plot_resid_fit(model, print_plot = TRUE)

Arguments

model

An object of class lm.

print_plot

logical; if TRUE, prints the plot else returns a plot object.

Details

Characteristics of a well behaved residual vs fitted plot:

  • The residuals spread randomly around the 0 line indicating that the relationship is linear.

  • The residuals form an approximate horizontal band around the 0 line indicating homogeneity of error variance.

  • No one residual is visibly away from the random pattern of the residuals indicating that there are no outliers.

See also

Other residual diagnostics: ols_plot_resid_box(), ols_plot_resid_hist(), ols_plot_resid_qq(), ols_test_correlation(), ols_test_normality()

Examples

model <- lm(mpg ~ disp + hp + wt, data = mtcars)
ols_plot_resid_fit(model)