Identification of the most underpredicted case

Source: R/deviant.R

The case with the largest positive difference between the observed value and the predicted value on the outcome. Depending on the research question, there might be a specific interest in the case for which the model performs worst and yields a smaller predicted value.

most_underpredicted(lmobject)

Arguments

lmobject

Object generated with lm

Value

The most underpredicted case with the largest positive residual (the most positive residual).

Examples

df <- lm(mpg ~ disp + wt, data = mtcars)
most_underpredicted(df)

#> Toyota Corolla 
#>        6.34844