This table is poorly structured for automating variable selection. For a given cluster solution the cluster identification variable Cluster: for "Cluster 1.... Cluster N" is only populated for the first variable in that cluster.  The remaining variables have a blank in Cluster until the first variable in the next Cluster.  Can anyone share a prog to autopopulate all variables within a cluster with their cluster identifier? Current: NumberofClusters          Cluster           Variable 10                                   Cluster 1           Var1 10                                                            Var2 10                                                            Var3 10                                   Cluster 2           Var4 10                                                            Var5 10                                                            Var6 Change to: NumberofClusters          Cluster           Variable 10                                   Cluster 1           Var1 10                                   Cluster 1           Var2 10                                   Cluster 1          Var3 10                                   Cluster 2           Var4 10                                   Cluster 2          Var5 10                                   Cluster 2          Var6 ... View more

The intention of this code to get every permutation of copay onto one line for the 3 by variables listed within the macro.  (i.e. SubscriberCopay_Crowns, SpouseCopay_Crowns, etc... for 5 different relationships , for 20 types of services). I've written loops on many occasions but this one in particular is giving me an error as specified in the discussion title. proc sql;   select distinct translate(translate(trim(ValidRelationshipDescription),'_','/'),'_',' ')   into :Relationship1 - :Relationship5 from copay2; quit; proc sql;   select distinct _NAME_ into :Service1 - :Service20 from copay3; quit; %macro expand; data copay4;   set copay3;   by Groupnumber NetworkGroupID NetworkGroupIDDescription;    %do i = 1 %to 5;       %do k = 1 %to 20;     %if first.NetworkGroupID %then %do;      %if _NAME_ = "&&service&k.." %then %do;      &&Relationship&i.._&&Service&k.. = &&Relationship&i..;      %end;    %end; %end; run; %mend expand; %expand; dataset copay3 example line: Groupnumber, NetworkGroupID, NetworkGroupIDDescription, _NAME_, Any, Dependants_Only, Subscriber_Spouse, Subscriber_Only, Spouse_Dependants 00000-00001                  1                         All Networks                        Crown    50                    .                           .                                .                         . ... View more

I've been reading through the documentation on this procedure and I cannot explicitly find (or perhaps identify) the appropriate syntax to accomodate a specific sampling objective.  I am looking to generate a random sample without replacement on a dataset that contains claims adjudicated by different users.  Because some users process more claims than others, I want the sampling methodology to assign each user an equal probability of being chosen in the random sample  (I do not need a specific number [or proportion] of observations in the random sample from each user as many of the documentation examples show).  This will purge the bias related to claims examiners who process more claims having a higher chance of being "randomly" chosen because more of their claim lines exist on the dataset. I will provide a couple of arbitrary parameters so someone can illustrate the appropriate syntax. Claimlines: 1,000,000 Random Sample Size: 100 Variable identifying the claims examiner: User Thanks! ... View more

Chapter 15 of Cody/Smith's text illustrate several methods to make changes to all numeric and/or character variables in a data set.  What I'm looking to do is to datepart and format date9. on all datetime variables in my dataset.  What is the alias for all datetime variables in a dataset to reference in an array (i.e. all numeric: _NUMERIC_)? ... View more