partial.resid.plot {asbio} | R Documentation |
The function creates partial residual plots which help a user graphically determine the effect of a single predictor with respect to all other predictors in a multiple regression model.
partial.resid.plot(Y, X, smooth.span = 0.8)
Y |
A vector of quantitative responses. |
X |
A matrix of explanatory variables. |
smooth.span |
Degree of smoothing for smoothing line. |
Creates partial residual plots (see Kutner et al. 2002). Smoother lines from lowess
and linear fits from lm
are imposed over plots to help an investigator determine the effect of a particular X variable on Y with all other variables in the model. The function automatically inserts explanatory variable names on axes.
Returns p partial residual plots, where p = the number of explanatory variables.
Ken Aho
Kutner, M. H., Nachtsheim, C. J., Neter, J., and W. Li. (2005) Applied linear statistical models, 5th edition. McGraw-Hill, Boston.
Soil.C<-c(13,20,10,11,2,25,30,25,23) Soil.N<-c(1.2,2,1.5,1,0.3,2,3,2.7,2.5) Slope<-c(15,14,16,12,10,18,25,24,20) Aspect<-c(45,120,100,56,5,20,5,15,15) Y<-as.vector(c(20,30,10,15,5,45,60,55,45)) X<-as.matrix(cbind(Soil.C, Soil.N, Slope, Aspect)) par(mfrow=c(2,2),mar=c(5,4,1,1.5)) partial.resid.plot(Y,X)