Have a bunch of records as below: INJECT 35-40 UNITS INTO THE SKIN DAILY  INJECT 45-70 UNITS INTO THE SKIN DAILY. 10-12 UNITS AFTER BREAKFAST, LUNCH AND DINNER VIA INSULIN PUMP AROUND 100-150 UNITS DAILY   Created two additional variables  to place strings to the immediate LEFT ("LOWERNG") & RIGHT ( "UPPRNG" ) of where "-" is found in the string.  I've identified where the "-" is in the variable with the index command. Can be any position in the variable, usually near the beginning of the string, but not always.  Using                                                                                                                IF HYPHEN >0 THEN LOWERNG = substr(INSTRUCTIONS,HYPHEN,HYPHEN-5); - five characters to the LEFT IF HYPHEN >0 THEN UPPRNG = substr(INSTRUCTIONS,HYPHEN,HYPHEN+5); - five characters to the RIGHT intent is  to grab five characters to the immediate right ( "UPPRNG" ) of "-" and  five characters to the immediate left ("LOWERNG").  Or so I thought.   Here are some results: (HYPHEN is the position of the "-") INSTRUCTION HYPHEN UPPRNG LOWERNG INJECT 0-8 UNITS INTO THE SKIN 3 TIMES DAILY. 9 -8 UNIT -8 U INJECT 10-20 UNITS INTO THE SKIN DAILY. 10 -20 UNI -20 U INJECT 40-55 UNITS INTO THE SKIN DAILY.  10 -55 UNI -55 U   UPPERNG & LOWERNG are identical (more or less), both variables containing strings from the RIGHT of where the hyphen is. Not what I expected at all - below is what I thought I was going to receive:   INSTRUCTION HYPHEN LOWERNG UPPRNG INJECT 0-8 UNITS INTO THE SKIN 3 TIMES DAILY. 9 ECT 0 8 UN INJECT 10-20 UNITS INTO THE SKIN DAILY. 10 ECT 10 20 UN INJECT 40-55 UNITS INTO THE SKIN DAILY.  10 ECT 40 55 UN   I can scrub out any characters later with                                                                                              COMPRESS(UPPRNG,'0123456789','k') ;                                                                                              COMPRESS(LOWERNG,'0123456789','k') ;                                                                                                                 After all the characters are scrubbed out, am going to add LOWERNG & UPPRNG, divide by two, (will need to use a put command in there somewhere) so I get an average of the upper and lower ranges.  I would like to exclude the hyphen in either  LOWERNG or UPPRNG results.                                                      But I'm stuck.   Thanx.   ... View more

Have some dynamic data sets with different variables each run. Looking very similar to the below.                              Banana_             Cherry             Ice                  Lemon          Heath_Bar ID                           Split                  Sundae           Swirl                  Frost             Crunch XCDF1483                1                        0                   1                         0                     0 JKTR48392               0                       1                    0                          0                    1 OPRT76231              1                       1                    0                          0                    1 VXIQ87126                1                       0                   1                          0                    0 KTUE61824               0                       1                   0                         1                     1   This code appears to be the solution  - for an array with variables that are all dates, each variable a different month.                                                                                                                                           data want; set have; _t=put(0,rb8.); array t(*) months1-months6;/**group your months vars with a double dash*/ want = n(of t(*))-count(peekclong (addrlong(t[1]), 48),_t) ; drop _t; run;   My variables are not dates, and the name of the variables are quite different each time. There would be maybe twenty variables at most.                                                                                                                                                           Result for each ID would be:    XCDF1483             2 JKTR48392            2 OPRT76231           3 VXIQ87126            2 KTUE61824           3   How do I count - and store the total count of non-zero values/non-null values for each unique ID in the data set  as a new variable?   Not counting "0" or null.  Again, result is stored in the data set as new variable "Total", or something similar.   Didn't use any string examples, but there will be some strings. Note:  A variable is either formatted numeric (0 or 1, nothing else) or a char. Character variables do not have numbers, and numerics do not have characters. Longest string variable is 30 letters.   Thanx. ... View more

Have a couple million records with a string like "00 00 01 00 00 01 00 01 00 00 00 00 01 01 00 01 00 00 00 00 01" "01 00 01 00 00 00 01 00 00 00 00 00 00 01 00 01 00 00 00 00 00" String has a length of 56. Positions 2/4/6/8/10, etc are filled with either a 1 or most likely, 0. Positions 1/3/5/7/9, etc. are all zeros and don't matter. . My job is parse the string of each record every two positions (there are no spaces, that is just for clarification). If there is a 1 in position two that means increment var1 +1 If there is ALSO a 1 in position four, (don't care about leading "0"'s in position 1/3/5/9, etc.) increment var2 + 1. Each two positions represent a different var, 28 total vars. String length is ALWAYS 56. The entire 56 len string must be parsed every two characters. Potentially there could be 28 variables that have to be incremented in each pass of the string, (but not realistic, most likely there is only five or six) which could be found in any section of the string, beginning to end (as long as they are in position 2/4/6/8 up to 56, etc.) Using the first example string, results would be Var1  0 Var2  0 Var3  1 Var4  0 Var5 0 var6  1 var7  0 Var8  1 .. Var13  1 Var14 1 var15 0 Var16 1 Var21 1  (skipping a bunch of vars with 0's) Now we process the second string. Results are now: Var1 1  Var2 0 (still zero, nothing in that position so far). Var3 1+1 = 2  Var4 0 (still zero) Var5 0 (still zero) Var6 1 (from previous row) Var7 1  Var8 0 .. Var14 1 Var15 0 Var 16 1 Var 21 1   This is what my boss gave me: **if substr(BigString,2,1)='1' then var1+1;** OK. Fine. A) There are 27 more places to evaluate in the string. B) there are a couple million records. 28 nested if then do loops doesn't sound like an answer (all I could think of). At least not to me. Cut-and-pasted a lot of looping code as below, which check every two positions , but I get a finished product with twice as many records   as I started with in a table I can query, but I'm looking to store variables var1-28 to access them later. (I don't have any choice about this). Thanx.   data have; set sources; Obs = _n_; len = length(sourcerow); do i=1 to len by 2; value = substr(sourcerow, i, 2); if value gt '00' then output;  end; run; ... View more

Receiving datasets from a variety of clients on a Redhat Enterprise Server 7 (Maipo) Need to check if the required 15 fields are present in the dataset, if not  then send message back to calling BASH program that the dataset submitted doesn't "pass muster", so to speak.  No further error handling is required, such as what fields are missing,etc,  at this point.   Using simple create table proc sql to create a tmp dataset consisting of  fields 1 - 15 from the submitted dataset. If any one field of the fifteen is missing, PROC SQL throws an error,(obviously), although the  NOERRORSTOP optiion will allow the code to continue to execute, (to check any other datasets submitted along with the first one).  My boss REALLY doesn't want any errors in the log, even if I'm later checking to see if  tmp datasets 1,2,3, etc. exist.  If  just one of the tmp tables weren't created by the proc sql create query, the code sends an error message back to be collected by the BASH shell script. I'm thinking PROC SQL is not the answer to determine if a dataset has a full complement of variables (but not what is IN the variables).  If more variables are in the dataset than required, they will NOT be included, they just drop into the ether.  How else can I check multiple datasets for a full complement of variables?   Code is really simple - Proc SQL; create table DS1 as select field1 field2, field3,(etc). from dataset; quit;     THANX. ... View more

Want to create a dataset containing one variable - the SAS temporary working directory on a  RedHat 7.7 (Maipo) server.  Code below works fine, until I try to write the (looooong) working  directory string to a temporary dataset.   CODE: proc options option = work;  <<<gets temporary SAS work directory run; %put %sysfunc(getoption(work)); %let tmpSASwrk = %sysfunc(getoption(work)); %put &tmpSASwrk;   data working;  <<<create a SAS data set "Working" with just 1 variable - temp work directory call symput("workdir",&tmpSASwrk); run;   Although the working directory is correctly identified, I get the following error:   "/saswork/SAS_work28AD0000FF4A_someplace.over.therainwbow.com"  <<<correct working directory 22 780 ERROR: DATA STEP Component Object failure. Aborted during the COMPILATION phase. ERROR 22-322: Syntax error, expecting one of the following: a name, a quoted string, a numeric constant, a datetime constant, a missing value, INPUT, PUT. ERROR 780-185: Invalid object or method name.   How can I save the current (temporary) working directory string to a dataset with one variable?   ... View more