Optimal cutoff-point for model classification

[haoduonge]

Hi all,

I would like to have a follow-up question on the topic asked earlier and locked.

Can anyone help with details on the mothod suggested here?

I follow this instruction, get Sp, Se, Pos predictive, Neg predictive but do not see each level of continuous variable (for example absence day).

Thank you!

https://communities.sas.com/t5/Statistical-Procedures/determine-significant-cut-off-points/m-p/79628

[StatDave]

If you specify the PLOTS(ONLY)=ROC(ID=ID) option in the PROC LOGISTIC statement and include the desired labeling variables in the ID statement, then the displayed ROC plot will label the points on the ROC curve with those variable values.