ols_plot_resid_fit: Residual vs fitted plot

In cmlopera/olsrr: Tools for Building OLS Regression Models

Description

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)

Arguments

model

An object of class lm.

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.

Deprecated Function

ols_rvsp_plot() has been deprecated. Instead use ols_plot_resid_fit().

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)

cmlopera/olsrr documentation built on May 26, 2019, 10:34 a.m.