Hi,   I have a file with almost 1,000 variables. I want a single report that lists every var in my file, the # of missing values, and the number of non-missing values:   Varname       nmiss      n var1               100         0 var2                 0        100 var3               50         50 .... etc.   I usually use proc iml for this, but I received an error message about not having enough memory, so I am looking for an alternate methodology. I know that missingness for character and numeric variables can be displayed using PROC FREQ, but I really want the output all together instead of separate for each variable. Note that I have too many vars to list all of them by hand, and that I have both numeric and character variables.    Any help is much appreciated! ... View more

Hi,   I have two data files. Each one has an ID variable, var1-var7 (which are the same variables on each file), and then the first file has some additional variables not found on the other file. The ID variable is unique on both files, BUT on File1 can sometimes be missing. Some of the IDs match across files, but there are also IDs that are unique to each. If the ID is missing, var1 - var7 will be missing, but other vars can be populated. If the IDs match across both files, var1-var7 will only be populated on File2.    FILE1 ID   var1    var2    var3....var7    other_vars... 1                                                     xyz 2       d                    e          f           xyz .                                                      xyz   FILE2 ID   var1    var2    var3....var7    1       a         b          c      d 3       a         b          c      d   What I want is to combine the files such that: 1. All records from File1 and File2 are kept (even those with no ID) 2. If a record is in both files, the var1-var7 values come from File2 3. If a record is just in File1 or just in File2, all values should remain intact for all variables.    I considered using UPDATE to update File1 using File2 (as the update process is essentially what I want to do), but the missing ID variables seemed to be problematic when I tried. I also tried using a data step merge and then a PROC SQL join, but in both of those cases, either the File1-only or File2-only values were being overwritten. ... View more

Hi everyone,   I'm having a very strange problem. I have a program which refers to a zip file using a filename statement:   filename zip1 zip "&dir./&file..zip";   Then later, I try to read in a SAS file from within the zip file using the infile statement:   data _null_;       infile zip1(C:/dirname/filename.sas7bdat)         lrecl=256 recfm=F length=length eof=eof unbuf;      -other unrelated stuff-   run;   This works just fine. The issue is that I wanted to automatically generate the name of the SAS file. I set up a step to do this and it produces exactly the value I wanted (C:/dirname/filename.sas7bdat).    The issue is that the data _null_ step above will work if I use the value of the macro variable written out (C:/dirname/filename.sas7bdat), but will generate an error message if I substitute that with &macrovar. I tried putting " " around &macrovar (no help), and macrotizing just parts of the filepath (which worked fine, but I want to macrotize the whole thing).    The error message I get says that the SAS file does not exist within the zip file.    Any help is much appreciated! ... View more

Hi,   I have two text files. Each of them has a continuous numeric variable (the same variable across both, ContVar); one of them also has a categorical var. Basically, I want to create a format in which all values of ContVar in either text file are valid values for Var1 in my SAS dataset.   Textfile 1 ContVar   CategVar 12345         abc 23456         def 34567         ghi etc.   Textfile 2 Contvar 890123 874321 etc.   I know the simple but inefficient way to do this (1. Create SAS dataset from textfile 1, 2. Create SAS dataset from textfile2, 3. Stack them and then create the format file). I'm hoping there is a more efficient way.   When I thought I only needed the values from Textfile 1, this is the formatting step I came up with:   data want (keep=start label hlo fmtname); infile "path/Textfile1.txt" end=eof; input start CategVar $; retain fmtname '$fmt_name'; retain label 'VALID'; output; if eof; start=' '; label='NOT VALID'; hlo='O'; output; run;   proc format cntlin=want;run;   I'm hoping Textfile 2 can just somehow be incorporated into that (or similar) rather than the process I outlined above.   Any help is much appreciated. ... View more