1. Jessica Bennett, Advance America Barbara Ross, Flexshopper LLC PharmaSUG 2015 Paper #QT06

2. 2 Jessica Bennett Fluent in SQL Loves R Learning to embrace SAS Barbara Ross Fluent in SAS Pretty good with SQL Opened R once

3. 3 Production Database via SQL & R

4. Convert SQL to SAS 4 Production Database +

5. This should be easy… That’s what PROC SQL is for!

6. a) Concatenating strings SQL Code: SELECT FirstName + ' ' + LastName As FullName FROM Customers MySQL Code: SELECT CONCAT(FirstName, ' ', LastName) As FullName FROM Customers SAS Code - “||” Operator: SELECT FirstName || ' ' || LastName As FullName FROM Customers SAS Code - CATX Function: SELECT CATX(' ',FirstName,LastName) As FullName FROM Customers 6

7. b) Not equals shortcut SQL Code - !=: SELECT * FROM inputds WHERE convert(date, date_created) != '1-5-2014‘ SQL Code - <>: SELECT * FROM inputds WHERE convert(date, date_created) <> '1-5-2014' SAS Code: SELECT * FROM inputds WHERE date_created ^= '05jan2014'd; 7

8. Global macro variables available in SAS the whole session In SQL, variable is only available while query is processing SQL Code: DECLARE @FirstMonday date SET @FirstMonday = '1-5-2014' SELECT * FROM inputds_201411 WHERE date > @FirstMonday SAS Code: %LET FirstMonday='05jan2014'd; PROC SQL; CREATE TABLE tmp AS SELECT * FROM inputds_201401 WHERE date>=&FirstMonday; QUIT; 8

9. - Temporary tables must be dropped with SQL. - SAS stores tables in the temporary work library by default - #TMP causes SQL to hold the temp table in memory the whole session SQL Code: IF OBJECT_ID('tempdb..#alldata') IS NOT NULL DROP TABLE #alldata SELECT * INTO #alldata FROM inputds_201411 SAS Code: PROC SQL; CREATE TABLE alldata AS SELECT * FROM inputds_201411; QUIT; 9

10. - SQL global tables are only available while query is processing (saves memory) SQL Code: DECLARE @TEST TABLE ( First_Name varchar(12), Date_Created datetime ) INSERT INTO @TEST VALUES ('Jessica', '1-9-2015') ('Barbara', '1-5-2015') SELECT * FROM @TEST SAS Code: DATA _null_; SET inputds; RUN; 10

11. In example, inputds1 cust_id is character & inputds2 cust_id is numeric SQL Code: SELECT * FROM inputds1 a INNER JOIN inputds2 b on b.cust_id = a.cust_id SAS Code – Convert both to numeric: PROC SQL; CREATE TABLE new AS SELECT * FROM inputds1, inputds2 ON INPUT(inputds1.cust_id,best32.) = inputds2.cust_id; QUIT; 11

12. SQL SYNTAX:CONVERT(data_type, expression) SAS SYNTAX: PUT(source, format) INPUT(source, informat) In example, SSN is numeric & amount is character (“$34.35”) SQL Code: SELECT RIGHT(‘00000000’ + CONVERT(VARCHAR(9), SSN), 9) as SSN,CONVERT(numeric, REPLACE(amount, '$', '') ) as amount FROM inputds SAS Code: PROC SQL; CREATE TABLE new AS SELECT PUT(SSN,z9.) as SSN,INPUT(amount,dollar20.) as amount FROM inputds; QUIT; 12

13. “d” & “dt” not needed in SQL If declared as a datetime, SQL will default to 0:0:0 SQL Code: SELECT * FROM inputds WHERE Date_Created >= '1-5-2014' -or – SELECT * FROM inputds WHERE Date_Created >= '2014/08/01' SAS Code: PROC SQL; CREATE TABLE tmp AS SELECT * FROM inputds WHERE DATEPART(date_created) >= '05jan2014'd; QUIT; 13

14. SQL Syntax: DATEADD(interval, increment int, expression); DATEDIFF(interval, startingdate, ending_date); SAS Syntax: INTNX(interval, start-from, increment, ); INTCK(interval, start-date, end-date); SQL Code: SELECT DATEADD(week, 2, LastPayDate) as NextPayDate,DATEDIFF(day, LastPayDate, getdate()) as DaysSincePaid FROM inputds SAS Code: PROC SQL; CREATE TABLE tmp AS SELECT INTNX('week',LastPayDate,2,'end') as NextPayDate,INTCK('day',LastPayDate,today()) as DaysSincePaid FROM inputds; QUIT; 14

15. Allowed variable name lengths:  SAS - 32 characters  Oracle - 30 characters  SQL - 128 characters (116 characters in temporary tables) SAS will crop/rename your variables to conform to their naming conventions if accessing SQL through a Libname SAS does not support variable names that start with a number. Spaces allowed in SQL & SAS EG. To have SAS replace spaces with underscores, there is the VALIDVARNAME option SAS Code: SAS OPTIONS VALIDVARNAME=v7; 15

16. There’s more! See the paper for more examples 16

17. There’s more! See the paper for more examples Even then it’s not everything. 17 PROC SQL for SQL Die-Hards Part 2?

18.  Biggest difference is changing the way you think about creating tables.  SAS is a analytical tool so it creates and changes tables more easily. It also tends to “hold on” to information (like macros and tables).  SQL programs are designed to minimize memory usage (hence the create a table then append on practice and only storing while in the current query). 18

19. Name: Barbara Ross Organization: Flexshopper LLC City, State: Boca Raton, FL E-mail: bmharlan@gmail.com Web: www.bharlan.weebly.com Name: Jessica Bennett Organization: Advance America City, State: Spartanburg, SC E-mail: Mary.Jessica.Bennett@gmail.com 19 Download paper & presentation here