log transformation

[peiy5920]

Hi all,

I have a data set and a variable named PCR in it. I want to normalize this variable, but I am not sure how to change it. Can someone give me some suggestions. Thanks.

 

[Kurt_Bremser]

Pkease provide example data innzsable form (data step with datalines), and what you expect to get out of it.

[ed_sas_member · Accepted solution]

Hi @peiy5920 

Here are some suggestions:

- to standardize the data (center around the mean and divide by standard deviation), you can use PROC STANDARD for example.

- to normalize the data, you can also try transformations: log transformation (in a data step, you define a new variable as NEWVAR = log(NEW)), square root transformation  (in a data step, you define a new variable as NEWVAR = sqrt(NEW)), ... You can also try Box and Cox transformation using PROC TRANSREG with the option BOXCOX in the MODEL statement.

[peiy5920]

Thank you so much!

[ed_sas_member]

You’re welcome 😊