polyfitGC function

Polynomial Regression

Regression analysis (one numerical predictor variable) with simplified output. Wrapper function for lm in package stats.

polyfitGC(form,data=parent.frame(),degree=2,graph=TRUE,check=FALSE)

Arguments

• form: formula of form y~x, both variables numeric

• data: dataframe supplying y and x above. If one or more of the variables is not in data, then they will be searched for in the parent environment.

• degree: desired degree of polynomial (for degree 1 use lmgC)

• graph: Produce scatterplot with fitted polynomial.

• check: Asks to produce a lowess or gam curve with approximate 95

  fitted line wanders outside the band, then perhaps a linear fit is not appropriate.

Returns

A list of class "polyGC". Elements that may be queried include "s" (residual standard error) and "R^2" (unadjusted).

Examples

#To study the relationship between two numerical variables:
polyfitGC(mpg~wt,data=mtcars,degree=2,graph=TRUE)
#check the second-fdegree fit:
polyfitGC(mpg~wt,data=mtcars,degree=2,check=TRUE)

Author(s)

Homer White hwhite0@georgetowncollege.edu