It's not clear whether the program is correct or not, but I can tell you what it does.  It is considering whether or not to add this to the SAS code: AND P.type_C = &type So you need to now the value of &TYPE to determine what happens. If the program is intending to check whether &TYPE has been assigned any value at all, it is incorrect.  Checking for >' ' is not the way to do it.  But it may be correct, it's just not possible to tell the intention of the programmer. ... View more

These solutions started me thinking ... would there be any value to SAS implementing column pointers for each row of the input buffer?  For example, could this approach be made to work with some tweaks to the software? data want; infile rawdata stopover; input #1 year #2 income @@; run; And would this be a tiny step toward implementing some functionality of Python or R within the SAS software? ... View more

To get your first DATA step to work, add a MISSING statement: data march; missing p k; input id name$ score; datalines; use the same data you are using now ; This lets you store p and k as numeric values without needing to change anything to 98. In fact, changing them to 98 would cause problems later. For example, when you calculate the average score, values of 98 would affect the average. ... View more

IF. Not WHERE. ... View more

You need IF, not WHERE. The IF statement can be used by any DATA step. The WHERE statement requires that the incoming data is already a SAS data set. Since your source of incoming data is text and not a SAS data set, you cannot use WHERE. ... View more

Here's a variation to take care of that:   data want; set have; by id date; retain flag 0; if first.date then flag = (last.date=0); run; ... View more

It looks like your data is already sorted by ID.  In that case, add a few more variables to complete the computations (then get rid of the ones not needed at the end): data want; set have; by id; if first.id then do; flagged_fd=0; flagged_reporting=0; end; retain flagged_fd flagged_reporting; newv_for_fd=0; newv_for_reporting=0; if fd then do; if flagged_fd=0 then newv_for_fd=1; flagged_fd=1; end; if reporting then do; if flagged_reporting=0 then newv_for_reporting=1; flagged_reporting=1; end; drop flagged_fd flagged_reporting; run; It's convoluted.  It's untested.  But it looks like it should work. ... View more

You can create an array holding only the variables that need to be converted.  This suggestion comes close but may include extra variables:   array charvals {*} _character_; That array might also include character variables from other parts of your data set, character variables that you may not want to convert.  A more accurate array statement would be: array chars {*} s09912_1-character-s16015_1; Note that you still need new variable names to hold the numeric result of your conversion.  An existing variable cannot be changed from character to numeric.  So there is more work to be done to accurately define those new names.  It would help if you show us how you plan to convert a single variable from character to numeric. ... View more

This statement doesn't compare anything to the string GE. In macro language, GE means "greater than or equal to". The easiest workaround would be to add double quotes to the comparison: %if "&brand" = "GE" %then %put Brand is General Electric; ... View more

Why should Company 1234 have a 1?  The change occurs in year 3, not in year 2.   What if there is more than one change (see below)?  What would you want as the result here?   Company ID     Year     manager_fem 1234                   1              0 1234                   2              0 1234                   3              1 1234                   4              1 4567                   1              0 4567                   2              1 4567                   3              0 4567                   4              1 ... View more

As long as parentheses are not the first character, you can use: inside = scan(variable, 2, ')('); This selects the second set of characters within the string, using parentheses as delimiters. ... View more

Here are a few possibilities to consider, before deciding on a solution.  Would the values below: stay as is change to "NA" doesn't matter, it won't happen 1 3 5 /* embedded blanks */ 3.1415 /* decimal point */ 246 /* leading blanks */ -789 /* negative sign */ +987 /* positive sign */ + 246 /* combinations */ If these are too easy, you might want to consider scientific notation such as  314159E-5 ... View more

Here's a program that produces the results in your post. data want; do until (last.id or indicaotr ne 0); set have; by id notsorted; end; final_indicaotr = indicaotr; do until (last.id or indicaotr ne 0); set have; by id notsorted; output; end; run; I hope I picked out the right result from your example.  Try it and see. ... View more

The format of the variables does not influence the outcome.  However ...   There is no way on God's green earth that those data sets combined to produce the output you have shown.  Neither data set contains an observation where ID has a missing value.  So the combined data set cannot contain such a value.     As a best guess, one data set contains slightly different values than posted.  For example, Table2 does not contain an observation with ID=1.  Instead, it contains an observation with ID=1.00000000001 and that is rounded to 1 for printing purposes.  That wouldn't explain everything about the results you got, but it would explain a lot.   Or perhaps ID is a character variable, and a couple of values in Table2 contain a trailing tab character instead of a trailing blank. ... View more