Hello,
How do you deal with a dependent variable that has some values equal to zero when using boxcox transformations in proc transreg?
proc transreg data = sample1;
model boxcox(DV)=identity(IV);
run;
ERROR: 11 invalid values were encountered while attempting to transform variable DV.
Thank you
You could prevent λ in the Box-Cox transformation from being exactly equal to zero, then you shouldn't have any problems with zero values of Y.
Thanks for your answer. I tried this and didn't work:
proc transreg data = sample1;
model boxcox(DV / lambda ^= 0) = identity(IV);
run;
proc transreg data = sample1;
model boxcox(DV / lambda = -3 to -0.25 by 0.25,-0.1, 0.1, 0.25 to 3 by 0.25) = identity(IV);
run;
Thanks for your response. I tried using the code with that specific lambda but it showed the same error.
Available on demand!
Missed SAS Innovate Las Vegas? Watch all the action for free! View the keynotes, general sessions and 22 breakouts on demand.
Learn the difference between classical and Bayesian statistical approaches and see a few PROC examples to perform Bayesian analysis in this video.
Find more tutorials on the SAS Users YouTube channel.