I am using proc tabulate to generate a table containing mean and standard deviation for some variables. It returns a table where the values are given with two significant figures. Is there any way I can round off these values to the nearest whole number instead?
Code used:
proc tabulate data=sved.rank4 missing order=formatted;
where MMSEny ne .;
class diagnosid overall;
var diagnosage MMSEny ab42ny ttauny ptauny;
tables diagnosage MMSEny ab42ny ttauny ptauny,
diagnosid=''*(mean stddev)
overall=''*(mean stddev);
format diagnosid diagnosid_. overall overall_.;
run
;
/Christoffer
proc tabulate data=sved.rank4 missing order=formatted;
where MMSEny ne .;
class diagnosid overall;
var diagnosage MMSEny ab42ny ttauny ptauny;
tables diagnosage MMSEny ab42ny ttauny ptauny,
diagnosid=''*(mean*f=best8.0 stddev)
overall=''*(mean stddev);
format diagnosid diagnosid_. overall overall_.;
run
The best format will show as most possible digits.
Do you want to round off at your wishes use a proper format for that according your wish
You don't mention the range of values so I'm guessing as to how large to make the format. If you need more digits increase the 6 to the number needed, up to 32 any way.
diagnosid=''*(mean*f=f6.0 stddev*f=f6.0)
overall=''*(mean*f=f6.0 stddev*f=f6.0);
It's finally time to hack! Remember to visit the SAS Hacker's Hub regularly for news and updates.
Learn how use the CAT functions in SAS to join values from multiple variables into a single value.
Find more tutorials on the SAS Users YouTube channel.
Ready to level-up your skills? Choose your own adventure.