Yes, Thank you. Totally understood. Next time I will approach the issue with my real problem, because if not, we will get confuse each other and the rest of the people want to learn from my issue. Thank you  very much again. V ... View more

hi ... yes you can get the same statistics with MEANS as you do with UNIVARIATE however, if you use UNIVARIATE and ODS OUTPUT to get the statistics into a SAS data set rather than the PROC option (OUTPUT OUT= ...) in either MEANS or UNIVARIATE, the data set orientation is different, for example ... ods listing close; ods output moments=ustat1 (keep=varname sex label1 nvalue1 where=(label1 in : ('N' 'Mean' 'Std'))); proc univariate data=sashelp.class; var height weight; class sex; run; ods output close; ods listing; gives you ...         Var Obs     Name     Sex    Label1                nValue1   1    Height     F     N                    9.000000   2    Height     F     Mean                60.588889   3    Height     F     Std Deviation        5.018328   4    Height     M     N                   10.000000   5    Height     M     Mean                63.910000   6    Height     M     Std Deviation        4.937937   7    Weight     F     N                    9.000000   8    Weight     F     Mean                90.111111   9    Weight     F     Std Deviation       19.383914 10    Weight     M     N                   10.000000 11    Weight     M     Mean               108.950000 12    Weight     M     Std Deviation       22.727186 while the follwoing ... proc univariate data=sashelp.class noprint; var height weight; class sex; output out=ustat2 n=n_ht n_wt mean=mean_ht mean_wt std=std_ht std_wt; run;    gives you ...   Obs Sex n_ht n_wt mean_ht mean_wt std_ht std_wt 1 F 9 9 60.5889 90.111 5.01833 19.3839 2 M 10 10 63.9100 108.950 4.93794 22.7272 (similar to what you would get with PROC MEANS output) the output you want depends on what comes next, what will you do with the data set (to get the output you wanted, I found it easier to use the ODS style output) one distinct advantage of the ODS output is that you do not have to provide all those variable names when using more tha one varaible in the VAR statement ... View more

Since you will surely have SUBJECTS or other grouping you will want INIT TO MISSING when a new subject is started to handle the possibility that a subject has no baseline.  ... View more

Thanks Astounding for your help, I will keep your piece of code in my mind, as well as the execute function proposed by Ksharp (I dont need it at the moment). Only just, I needed to be more "efficient" (I known is not the right word) format, as Ksharp suggested. ... View more

thnaks, I did with n ... View more

data_null...in termns of using proc means instead of proc summary...is not possible to get the same result...can you give me some suggestion of the formatting using proc means?...Thanks. ... View more

En...... 1) As you said, CLASS don't need to pre-sort. But BY is more efficient than CLASS, especially when you have a larget dataset. 2)There is no little different between MEANS and SUMMARY. SUMMARY is usually used when you only want to generate a output dataset. MEANS is usually to display the output in the LOG, if you only need to see the result ,then MEANS is a good choice. Ksharp ... View more

One or more is new feature in SAS 9.2. ... View more

As Cynthia said .STAT forum is your better place to find answer. You should choose the eigenvetor whose eigenvalue is great than 1, Maybe you can get two or three eigenvetor, it is dependent on your correlative matrix. Ksharp ... View more

Hi guys, I dont know about using the proc qsl for the structure I wanted: proc sort nodupkey proc frequency proc transpose I understand it looks like more code programming, but it is more intuitive to me Thanks anyway for the discussion. ... View more