BookmarkSubscribeRSS Feed
mjheever
Obsidian | Level 7

Hi everyone,

 

This is more stats-related question than a technical SAS-coding question, but maybe there is an option in the proc logistic procedure to generate the output I'm looking for.

 

I used a logistic regression model, proc logistic procedure, to model the probability of a binary event occuring (=1):

Event_ID Probability/ Expected Actual/ Event_Occured
123 0.98 1
223 0.67 1
224 0.012 1
225 0.0506 1
321 0.0006 0
322 0.103 0
323 0.0089 0
342 0.46 1
564 0.69 1
597 0.89 1
598 0.76 1
601 0.0045 0
689 0.15 0
756 0.68 1
897 0.72 1
899 0.089 0
1001 0.23 0
1501 0.88 1
8997 0.45 1

 

Is there an efficient way to graph the expected (predicted probabilities) vs actual (event results) to give an indication on how well the model is performing ?

 

Thank you !

1 REPLY 1
Rick_SAS
SAS Super FREQ

Yes, use the EFFECTPLOT statement:

ods graphics on;
proc logistic data=sashelp.class;
model sex = height;
effectplot;
run;

sas-innovate-2024.png

Available on demand!

Missed SAS Innovate Las Vegas? Watch all the action for free! View the keynotes, general sessions and 22 breakouts on demand.

 

Register now!

What is ANOVA?

ANOVA, or Analysis Of Variance, is used to compare the averages or means of two or more populations to better understand how they differ. Watch this tutorial for more.

Find more tutorials on the SAS Users YouTube channel.

Discussion stats
  • 1 reply
  • 1310 views
  • 0 likes
  • 2 in conversation