topic Re: Regression Plot with Confidence 95% CI Bounds in Graphics Programming

Regression Plot with Confidence 95% CI Bounds

ugoagwup — Thu, 10 Jun 2021 14:07:24 GMT

Hello!

I need some assistance. A simple linear regression script such as the following would give me a regression plot with 95% CI bounds on defaultRegression Plot.

Proc glm data= Tdata;

model LFTvalues = bckTrans_auc/ solution;
run;

Please how can I use Proc SGplot to replicate exact same plot and have full control of things display like removing the title from the plot?

Re: Regression Plot with Confidence 95% CI Bounds

Reeza — Thu, 10 Jun 2021 14:12:49 GMT

Follow the instructions here: https://blogs.sas.com/content/iml/2012/08/01/data-fro-ods-graphics.html

Re: Regression Plot with Confidence 95% CI Bounds

ugoagwup — Thu, 10 Jun 2021 14:33:29 GMT

Already figured out a way to solve my problem

proc sgplot data= Tdata;

reg x=bckTrans_auc y=LFTvalues/CLM CLI;

yaxis label = "AST (Units per liter)";
xaxis label = "AUC(h*ng/mL)";

Run;

Re: Regression Plot with Confidence 95% CI Bounds

ugoagwup — Thu, 10 Jun 2021 14:53:01 GMT

Thanks @Reeza

The following codes solved my problem

proc sgplot data= Tdata; reg x= bckTrans_auc y=LFTvalues/CLM CLI; yaxis label = "AST (Units per liter)"; xaxis label = "AUC(h*ng/mL)"; Run;

Please do you know how I can include the regression equation on the plot?

Re: Regression Plot with Confidence 95% CI Bounds

Reeza — Thu, 10 Jun 2021 14:56:46 GMT

AFAIK there isn't an option for that but you can check the INSET statement or REG statement options which is where I'd expect to see it.
Otherwise, here's an alternative method.
https://blogs.sas.com/content/iml/2013/02/27/slope-of-a-regression-line.html