Fluorite | Level 6

Visualizing Restricted Cubic Spline Effect for Multivariable Linear Model

Hello,

I am running a multivariable linear model with restricted cubic splines for my main exposure of interest. I am having difficulty trying to graph the predicted curve of my model, as the predicted value appears very jagged. I have referenced the following SAS post 'Visualize multivariate regression models by slicing continuous variables' (https://blogs.sas.com/content/iml/2017/12/18/visualize-multivariate-regression-models-by-slicing-con...), however to no avail.

Is anyone able to share guidance to help with my visualization? My code is as follows:

``````ODS SELECT ANOVA ParameterEstimates SplineKnots;
EFFECT spl = spline(exp1 / DETAILS naturalcubic BASIS=tpf(noint)
KNOTMETHOD=percentilelist(5 27.5 50 72.5 95));
CLASS sex city income;
MODEL IQ = spl sex city income weight/ SELECTION=none; /* fit model by using spline effects */
OUTPUT OUT=SplineOut PREDICTED=Fit;   /* output predicted values for graphing */
QUIT;

TITLE "Restricted Cubic Spline Regression";
PROC SGPLOT DATA=SplineOut NOAUTOLEGEND;
SCATTER x=exp1 y=IQ;
SERIES x=exp1 y=Fit / LINEATTRS=(THICKNESS=3 COLOR=red);
RUN;``````
3 REPLIES 3
Barite | Level 11

Re: Visualizing Restricted Cubic Spline Effect for Multivariable Linear Model

R is a free software that may show it nicer: https://search.r-project.org/CRAN/refmans/Hmisc/html/rcspline.plot.html
If you install R, their forum is this:
https://community.rstudio.com/
Fluorite | Level 6

Re: Visualizing Restricted Cubic Spline Effect for Multivariable Linear Model

Thank you for the recommendation.

Super User

Re: Visualizing Restricted Cubic Spline Effect for Multivariable Linear Model

Discussion stats
• 3 replies
• 365 views
• 1 like
• 3 in conversation