We help IT Professionals succeed at work.

T-SQL--MAX Is Not Working III:  The Search for More Knowledge

TBSupport asked
Last Modified: 2014-05-07

Sigh.....I thought that I had this cleared up in a posting of mine from yesterday.  But, I don't.  For goodness sake, I want one flippin' TRXBEGDT record.  That's not what I'm getting, when I use MAX.

Once again, please review the query and let me know what to do for goodness sake.

I don't want to hear anything about rewrites, CTEs, new inner joins, etc.


SELECT    'MED' AS company, CAST('202EMPLPIM' AS CHAR(15)) AS rectype, CAST('16860' AS CHAR(16)) AS cocode, CAST(MED.dbo.UPR00100.SOCSCNUM AS CHAR(11)) 
                      AS ssn, CAST(MED.dbo.UPR00100.EMPLOYID AS CHAR(64)) AS empl_id, CAST(MED.dbo.UPR00100.SOCSCNUM AS CHAR(64)) AS user_id, CAST('A868' AS CHAR(4)) 
                      AS inhousenum, CONVERT(VARCHAR(8), MAX(MED.dbo.UPR00900.LSTPCKDT), 112) AS pasofdate, CAST(MED.dbo.UPR00100.FRSTNAME AS CHAR(64)) AS fn, 
                      CAST(MED.dbo.UPR00100.MIDLNAME AS CHAR(64)) AS mn, CAST(MED.dbo.UPR00100.LASTNAME AS CHAR(64)) AS ln, CAST('' AS CHAR(12)) AS suffix, 
                      CAST('' AS CHAR(5)) AS title, CAST(RTRIM(CAST(RIGHT(MED.dbo.UPR00100.SOCSCNUM, 10) AS VARCHAR(10))) + CAST(YEAR(MED.dbo.UPR00100.BRTHDATE) 
                      AS VARCHAR(4)) AS CHAR(8)) AS dfpin, CAST('Y' AS CHAR(1)) AS direct_login, CAST(MED.dbo.UPR00100.LOCATNID AS CHAR(12)) AS verdiv, 
                      CAST(MED.dbo.UPR00100.JOBTITLE AS CHAR(64)) AS jobtitle, CAST((CASE UPR00100.LASTDAYWORKED_I WHEN '1/1/1900' THEN 'A' ELSE 'G' END) AS CHAR(1)) 
                      AS eestatcd, CONVERT(VARCHAR(8), MED.dbo.UPR00100.STRTDATE, 112) AS mrhdate, CAST('' AS CHAR(1)) AS eestattype, CONVERT(VARCHAR(8), 
                      MED.dbo.UPR00100.STRTDATE, 112) AS orighiredate, CAST('' AS CHAR(8)) AS termdate, CAST('' AS CHAR(2)) AS termreason, CAST('9304' AS CHAR(10)) 
                      AS uctermreason, CAST(CASE LASTDAYWORKED_I WHEN '19000101' THEN '' ELSE CONVERT(VARCHAR(8), MED.dbo.UPR00100.LASTDAYWORKED_I, 112) 
                      END AS CHAR(8)) AS lastdaywrkf, CAST(MED.dbo.UPR00102.STATE AS CHAR(2)) AS wrkstate, CAST('100' AS CHAR(20)) AS wrkloccd, CAST('04249741' AS CHAR(15)) 
                      AS fein, CAST('09363930' AS CHAR(15)) AS suiacctnum, CONVERT(VARCHAR(8), MED.dbo.UPR00100.STRTDATE, 112) AS adjhiredate, CAST('000' AS CHAR(3)) 
                      AS yrsofserv, CAST('00' AS CHAR(2)) AS mthofserv, 
                      CAST((CASE WHEN UPR00400.PAYUNPER = 1 THEN '07' WHEN UPR00400.PAYUNPER = 2 THEN '06' WHEN UPR00400.PAYUNPER = 3 THEN '05' WHEN UPR00400.PAYUNPER
                       = 4 THEN '04' WHEN UPR00400.PAYUNPER = 5 THEN '03' WHEN UPR00400.PAYUNPER = 6 THEN '02' WHEN UPR00400.PAYUNPER = 7 THEN '01' WHEN UPR00400.PAYUNPER
                       = 8 THEN '08' END) AS CHAR(2)) AS payfreq, CAST('' AS CHAR(1)) AS twn_add, CAST('' AS CHAR(1)) AS ucx_add, CAST('' AS CHAR(1)) AS ignore_bp, 
                      CAST('U' AS CHAR(1)) AS add_ind_1, CAST('home' AS CHAR(20)) AS type_1, CAST(MED.dbo.UPR00102.ADDRESS1 AS CHAR(60)) AS I1_1, 
                      COALESCE (CAST(MED.dbo.UPR00102.ADDRESS2 AS CHAR(60)), '') AS I2_1, COALESCE (CAST(MED.dbo.UPR00102.ADDRESS3 AS CHAR(60)), '') AS I3_1, 
                      CAST(MED.dbo.UPR00102.CITY AS CHAR(60)) AS city_1, CAST(MED.dbo.UPR00102.STATE AS CHAR(2)) AS state_1, CAST(MED.dbo.UPR00102.ZIPCODE AS CHAR(16)) 
                      AS zip_1, CAST('US' AS CHAR(2)) AS country_1, CAST('' AS CHAR(20)) AS county_1, CAST('' AS CHAR(64)) AS province_1, CAST('' AS CHAR(1)) AS add_ind_2, 
                      CAST('' AS CHAR(20)) AS type_2, CAST('' AS CHAR(60)) AS I1_2, CAST('' AS CHAR(60)) AS I2_2, CAST('' AS CHAR(60)) AS I3_2, CAST('' AS CHAR(60)) AS city_2, 
                      CAST('' AS CHAR(2)) AS state_2, CAST('' AS CHAR(16)) AS zip_2, CAST('' AS CHAR(2)) AS country_2, CAST('' AS CHAR(20)) AS county_2, CAST('' AS CHAR(64)) 
                      AS province_2, CAST('' AS CHAR(1)) AS add_ind_3, CAST('' AS CHAR(20)) AS type_3, CAST('' AS CHAR(60)) AS I1_3, CAST('' AS CHAR(60)) AS I2_3, CAST('' AS CHAR(60)) 
                      AS I3_3, CAST('' AS CHAR(60)) AS city_3, CAST('' AS CHAR(2)) AS state_3, CAST('' AS CHAR(16)) AS zip_3, CAST('' AS CHAR(2)) AS country_3, CAST('' AS CHAR(20)) 
                      AS county_3, CAST('' AS CHAR(64)) AS province_3, CAST('' AS CHAR(1)) AS add_ind_4, CAST('' AS CHAR(20)) AS type_4, CAST('' AS CHAR(60)) AS I1_4, 
                      CAST('' AS CHAR(60)) AS I2_4, CAST('' AS CHAR(60)) AS I3_4, CAST('' AS CHAR(60)) AS city_4, CAST('' AS CHAR(2)) AS state_4, CAST('' AS CHAR(16)) AS zip_4, 
                      CAST('' AS CHAR(2)) AS country_4, CAST('' AS CHAR(20)) AS county_4, CAST('' AS CHAR(64)) AS province_4, CAST('' AS CHAR(1)) AS home_phone_ind, 
                      CAST('' AS CHAR(3)) AS home_phone_country, CAST('' AS CHAR(16)) AS home_phone, CAST('' AS CHAR(1)) AS work_phone_ind, CAST('' AS CHAR(3)) 
                      AS work_phone_country, CAST('' AS CHAR(16)) AS work_phone, CAST('' AS CHAR(6)) AS work_phone_ext, CAST('' AS CHAR(1)) AS cell_phone_ind, CAST('' AS CHAR(3)) 
                      AS cell_phone_country, CAST('' AS CHAR(16)) AS cell_phone, CAST('' AS CHAR(1)) AS pager_ind, CAST('' AS CHAR(3)) AS pager_country, CAST('' AS CHAR(16)) AS pager, 
                      CAST('' AS CHAR(6)) AS pager_pin, CAST('' AS CHAR(1)) AS fax_ind, CAST('' AS CHAR(3)) AS fax_country, CAST('' AS CHAR(16)) AS fax, CAST('' AS CHAR(6)) AS fax_ext, 
                      CAST('' AS CHAR(1)) AS other_ind, CAST('' AS CHAR(3)) AS other_phone_country, CAST('' AS CHAR(16)) AS other_phone, CAST('' AS CHAR(6)) AS other_phone_ext, 
                      CAST('' AS CHAR(256)) AS email1, CAST('' AS CHAR(256)) AS email2, CAST('' AS CHAR(256)) AS email3, CAST('' AS CHAR(256)) AS email4, CAST('' AS CHAR(256)) 
                      AS email5, CAST('' AS CHAR(256)) AS email6, CAST((CASE UPR00100.MARITALSTATUS WHEN '1' THEN 'M' WHEN '2' THEN 'S' ELSE 'O' END) AS CHAR(1)) 
                      AS marital_status, CAST((CASE UPR00100.GENDER WHEN '1' THEN 'M' WHEN '2' THEN 'F' ELSE 'U' END) AS CHAR(1)) AS gender, 
                      CAST(CASE UPR00100.BRTHDATE WHEN '1/1/1900' THEN '' ELSE CONVERT(VARCHAR(8), UPR00100.BRTHDATE, 112) END AS CHAR(8)) AS birth_date, 
                      CAST('' AS CHAR(256)) AS employer_def_1, CAST('' AS CHAR(256)) AS employer_def_2, CAST('' AS CHAR(256)) AS employer_def_3, CAST('' AS CHAR(256)) 
                      AS employer_def_4, CAST('' AS CHAR(1)) AS employer_log_def_1, CAST('' AS CHAR(1)) AS employer_log_def_2, CAST('' AS CHAR(1)) AS employer_log_def_3, 
                      CAST('' AS CHAR(1)) AS employer_log_def_4, CAST('' AS CHAR(1)) AS filler2, CAST('I' AS CHAR(1)) AS consent_flag, CAST('00000000' AS CHAR(8)) AS consent_date, 
                      CAST('' AS CHAR(1)) AS primary_email, CAST('' AS CHAR(1)) AS secondary_email, CAST('' AS CHAR(7)) AS address, CAST('' AS CHAR(4)) AS daytime_phone, 
                      CAST('' AS CHAR(4)) AS evening_phone, CAST('' AS CHAR(73)) AS filler, CAST('71912' AS CHAR(5)) AS tci_id, CAST('' AS CHAR(1)) AS wotc_add, CAST('' AS CHAR(1)) 
                      AS tics_add, CAST('' AS CHAR(32)) AS tax_group_name, MAX(CONVERT(VARCHAR(8), MED.dbo.UPR30300.TRXBEGDT, 112)) AS pay_process_begin_date, 
                      CAST('' AS CHAR(1)) AS filler1, CAST('' AS CHAR(1)) AS filler8, CAST('' AS CHAR(1)) AS filler3, CAST('' AS CHAR(1)) AS filler4, CAST('' AS CHAR(1)) AS filler5, 
                      CAST('' AS CHAR(1)) AS filler6, CAST('' AS CHAR(1)) AS filler7, CAST('' AS CHAR(1)) AS Expr1, CAST('' AS CHAR(1)) AS filler9, CAST('' AS CHAR(143)) AS filler10
FROM         MED.dbo.UPR30300 INNER JOIN
                      MED.dbo.UPR00900 ON MED.dbo.UPR30300.CHEKNMBR = MED.dbo.UPR00900.LPCHKNUM AND 
                      MED.dbo.UPR30300.EMPLOYID = MED.dbo.UPR00900.EMPLOYID INNER JOIN
                      MED.dbo.UPR00100 ON MED.dbo.UPR30300.EMPLOYID = MED.dbo.UPR00100.EMPLOYID INNER JOIN
                      MED.dbo.UPR00102 ON MED.dbo.UPR00100.EMPLOYID = MED.dbo.UPR00102.EMPLOYID INNER JOIN
                      MED.dbo.UPR00400 ON MED.dbo.UPR00100.EMPLOYID = MED.dbo.UPR00400.EMPLOYID
WHERE     (MED.dbo.UPR30300.UPRTRXAM > 0) and MED.dbo.UPR00100.SOCSCNUM = '018786594'
                      MED.dbo.UPR00100.MIDLNAME, MED.dbo.UPR00100.LASTNAME, MED.dbo.UPR00100.JOBTITLE, MED.dbo.UPR00100.STRTDATE, 
                      MED.dbo.UPR00100.BENADJDATE, MED.dbo.UPR00400.PAYUNPER, MED.dbo.UPR00102.ADDRESS1, MED.dbo.UPR00102.ADDRESS2, 
                      MED.dbo.UPR00102.ADDRESS3, MED.dbo.UPR00102.CITY, MED.dbo.UPR00102.STATE, MED.dbo.UPR00102.ZIPCODE, MED.dbo.UPR00102.PHONE1, 
                      MED.dbo.UPR00100.MARITALSTATUS, MED.dbo.UPR00100.GENDER, MED.dbo.UPR00100.BRTHDATE, MED.dbo.UPR00100.LASTDAYWORKED_I, 
                      MED.dbo.UPR30300.PAYROLCD, MED.dbo.UPR30300.PAYRATE, MED.dbo.UPR00100.LOCATNID

Open in new window

Watch Question

Kent OlsenData Warehouse / Database Architect

Hi TB,

This goes back to the point I made in an earlier post today.  And your challenge that "don't want to hear anything about rewrites, CTEs, new inner joins, etc." is counter productive.

The MAX value is based on an identifiable set of rules.  Your query needs to find the MAX value based on those rules, not a set of rules that work for this one specific example.  Those kinds of queries fail as soon as the data changes and you're right back debugging SQL all over again.

Let's start with the basics.  What are the rules for determining MAX?  Is it by company?  SSN?  SSN and Date?  



I just want the maximum TRXBEGDT for each ssn.  I don't know what else to describe.
Kent OlsenData Warehouse / Database Architect

Then the query structure is pretty easy.  :)

SELECT ssn, MAX (TRXBEGDT) maxdt FROM sometable  t0
  SELECT all of the other detail
} t1
  ON t0.ssn = t1.ssn
 AND t0.maxdt = t1.TRXBEGDT

Open in new window

There are other forms that work equally well, but that's the logic behind what you need to do.



I don't understand.


This is ridiculous.  If you have to go to all of this trouble just to get a maximum value, then Microsoft needs to revamp it's T-SQL books and online paraphernalia to say so.
EE Topic Advisor
Most Valuable Expert 2014
Awarded 2013
This one is on us!
(Get your first solution completely free - no credit card required)


Hi PortletPaul:

Thank you, for getting me past this!  I resolved this simply by placing MAX on every field in the select clause except ssn and only placed ssn in the GROUP BY section.

Again, thank you and thank you for putting up with my frustration!

Kent OlsenData Warehouse / Database Architect

Hi TB,

Just so you understand, all of those MAX values don't have to come from the same row.....
PortletPaulEE Topic Advisor
Most Valuable Expert 2014
Awarded 2013

Excellent news. Sorry it took us a while but you hadn't really told us much about the "expected result", but now we all know.

(most of the time)
 few items in  the group by = few result rows
 lots of items in the group by = lots of result rows

>>"all of those MAX values don't have to come from the same row"
then MAX() on the non-grouping fields is fine

Gain unlimited access to on-demand training courses with an Experts Exchange subscription.

Get Access
Why Experts Exchange?

Experts Exchange always has the answer, or at the least points me in the correct direction! It is like having another employee that is extremely experienced.

Jim Murphy
Programmer at Smart IT Solutions

When asked, what has been your best career decision?

Deciding to stick with EE.

Mohamed Asif
Technical Department Head

Being involved with EE helped me to grow personally and professionally.

Carl Webster
CTP, Sr Infrastructure Consultant
Empower Your Career
Did You Know?

We've partnered with two important charities to provide clean water and computer science education to those who need it most. READ MORE

Ask ANY Question

Connect with Certified Experts to gain insight and support on specific technology challenges including:

  • Troubleshooting
  • Research
  • Professional Opinions
Unlock the solution to this question.
Join our community and discover your potential

Experts Exchange is the only place where you can interact directly with leading experts in the technology field. Become a member today and access the collective knowledge of thousands of technology experts.

*This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.


Please enter a first name

Please enter a last name

8+ characters (letters, numbers, and a symbol)

By clicking, you agree to the Terms of Use and Privacy Policy.