About Tom

Tom · ‎08-31-2011

If you just want one observation you could dispense with the CARDS. You could use a RETAIN statement. Or if you want to initialize everything to the same value you could use an ARRAY statement and take advantage of the NN*value syntax of the initial values. data ulsound1; array _num cid aaa (2*0); run;

Tom · ‎08-31-2011

You could fully specify the valid range. '1','2','3','4','5','6','7','8','9' ,'10' - '19' ,'20' - '29' ,'30' - '39' .... = _same_

Tom · ‎08-31-2011

Normally to generate multiple datasets you will need to use some type of code generation tool. Either SAS macro code or you can do it yourself with a data step and a %INC. filename code temp; data _null_; set have ; by client_name ; if first.client_name; file code; where = quote(trim(client_name)); put 'data split_' client_name ';' / ' set have;' / ' where client_name=' where ';' / 'run;' ; run; %inc code / source2 ;

Tom · ‎08-31-2011

There are two main differences in the MERGE and the SQL FULL JOIN. One is the way to check which dataset has contributed data. The CASE example above is most similar to the IF/THEN logic in your DATA step. The other is what it does when there are variables with the same name in both datasets. For this you will need to use the COALESCE function. Remember to use for both your BY variables and any other other variables that occur in both datasets.

Tom · ‎08-31-2011

Just add it to your INFORMAT definition. You will need to add the quotes that you have omitted for the other values. proc format; invalue month_pncf (default=2) '0' = 88 '99','.','-' = .U other = _same_; run;

Tom · ‎08-30-2011

The EG programm will be submitted to a SAS server. Most likely the SAS server is not setup properly to access the DB2 database. If access to DB2 is anything like access to Oracle there are probably environment variables or search paths that need to be set in the unix shell before the server is launched.

Tom · ‎08-30-2011

The problem with TRANWRD is that it will replace all instances in the string. So X_STD_Y_STD would become X_Y. You might want to test that _STD are the last four characters of the name. if reverse('_STD') =: reverse(trim(name)) then ... or if scan(name,-1,'_') = 'STD' then

Tom · ‎08-30-2011

If your input fields are all numbers then you can use data step logic to do it without macro code. Create an array of the target variables and another with the expected length of each variable. Then loop over the arrays keeping your own pointer into the input buffer. You can adjust the length depending on whether the value starts with a minus sign. data out; array var f1-f6 ; array len (6) _temporary_ (6*4) ; input @; col=1; do _n_ = 1 to dim(var); neg=substr(_infile_,col,1)='-'; var(_n_)=input(substr(_infile_,col,len(_n_)+neg),best.); col+len(_n_)+neg; end; drop col neg ; put (var(*)) (best6.); cards; 0.1 0.2 0.3 0.4 0.5 0.6 OK 0.110.220.330.440.550.66 OK 0.11-0.220.330.440.550.66 f2 shifts by 1 0.11-0.220.33-0.440.550.66 f2, f4 each adds 1 0.11-0.22-0.33-0.44-0.55-0.66 f2, f3, f4, f5, f6 all adds 1 -0.11-0.22-0.33-0.44-0.55-0.66 everything is wider by 1 -0.11-0.220.33-0.440.55-0.66 ; 0.1 0.2 0.3 0.4 0.5 0.6 0.11 0.22 0.33 0.44 0.55 0.66 0.11 -0.22 0.33 0.44 0.55 0.66 0.11 -0.22 0.33 -0.44 0.55 0.66 0.11 -0.22 -0.33 -0.44 -0.55 -0.66 -0.11 -0.22 -0.33 -0.44 -0.55 -0.66 -0.11 -0.22 0.33 -0.44 0.55 -0.66

Tom · ‎08-29-2011

Or conversely, the problem with allowing SASUSER to be read-write is that it is where your SAS profile lives. This means that when you open a second SAS session you do not get your normal default environment. I normally run with the -RSASUSER option so that all of my SAS sessions start with the same settings. You can always run one with the -NORSASUSER option when you want to make any changes to your profile.

Tom · ‎08-29-2011

Problem would probably be easier for you to understand and for others to help you with if you could explain more about what you are actually doing. What do these files represent. Why do you only want the last value for each observation? Normally if you want to combine data from two different files there should be some key variables that you can use to match (merge) the records. In your example the first value of every line seems to be the same as the line number. Is that an ID variable? Didn't you receive any specifications on what the content of these files was?

Tom · ‎08-29-2011

Are you sure that you haven't defined the slash as the split character in your PROC REPORT statement? I frequently use SPLIT='00'x to avoid having proc report eat any of the real characters in my data or labels.

Tom · ‎08-29-2011

It sounds like you want to split your ids into connected subgraphs. If you have SAS/OR licensed I have had success using PROC NETDRAW to do this. Otherwise you can code your own logic. I would start by converting your data from the current four variables to two variables and populate it with pairs of values from your four variables. data relations; set test; array x pan1-pan3 add1 ; do i=1 to dim(x)-1; do j=i+1 to dim(x); id1=x(i); id2=x(j); output; id1=x(j); id2=x(i); output; end; end; keep id1 id2 ; run; proc sort data=relations nodupkey; by id1 id2; run; You can grow your own subnets by looping and adding in all of the connected ids from the current list until the current subnet stops growing.

Tom · ‎08-24-2011

It does not look like the psftp program does what you want it to do. You can test that by taking SAS out of the problem and issuing the proposed command from a DOS prompt. I think it would be much easier to use SAS to write the script for psftp than to depend on an external file. But you could use an external file as a model for the script that SAS would write. The datastep below will replace all references to '%1' with the actual directory. %let sdir=d:\raj ; filename script temp; data _null_; infile 'abc.bat' ; file script; input; _infile_ = tranwrd(_infile_,'%1',"&sdir"); put _infile_; run; %sysexec psftp.exe username@hostname -p portnumber(22) -pw password -batch -b %sysfunc(pathname(script)) ;

Tom · ‎08-21-2011

Not finding the correct values is probably a result of the lines in your files being longer than the default line length for an INFILE statement. You should see notes in the log to the effect that some lines were truncated. Add a LRECL option to your INFILE statement such as LRECL=10000. Art's program is taking advantage of the fact that SAS can support a single wildcard (*) in a filename. You can test if it finds the files you want by issuing a DIR command with the same filename at the command line prompt on your PC. The INPUT statement without any variables will force SAS to read the line into the input buffer. You can reference this buffer using the special variable name _INFILE_. The SCAN function with a negative argument will scan backwards from the end of the string. You might want to specify what delimiters to use when scanning. For example change this to SCAN(_INFILE_,-1,' ') to force it to use space as the delimiter when scanning. Your example file was delimited with spaces and not tabs as you said above. The INPUT function will convert the characters pulled off the end of the line into a number using the BEST12. format. If you want to save the name of the individual file that a particular record came from in the output dataset then you will need to assign it to another variable as the variable named in the FILENAME= option of the INFILE statement will be automatically dropped from the output dataset.

Tom · ‎08-20-2011

Make sure that you have permission to run X command on the DIS server. Try running another program that does a different command. For example: x "echo hello > %sysfunc(pathname(work))\hello.dat" ; data _null_; infile "%sysfunc(pathname(work))\hello.dat"; input; list; run; If that works then try to find out if you have permission to run the cscript command on the server.

Online Status	Offline
Date Last Visited	Tuesday