To_date function not fetching the desired result.....

Question

Hi Gurus,
I get varied results from the below 2 queries, I am pessimistic about using to_char instead prefer using to_date, but query doesn't seem to fetch the complete data while using the to_date but with to_char it works. Please can someone make this to_date used query below work for me.
Main table structure:
Column Name     ID     Pk     Null?     Data Type     Default     Histogram     Encryption Alg     Salt
MINISTRY_CODE     1          Y     VARCHAR2 (15 Byte)          Yes          
BILL_MONTH     2          Y     DATE          Yes          
CUBIC     3          Y     NUMBER          Yes          
GALLONS     4          Y     NUMBER          Yes          
AMOUNTS     5          Y     NUMBER          Yes          
MTR_TYPE     6          Y     CHAR (1 Byte)          Yes          1st query:
SELECT ministry_code,
DECODE(mtr_type, 'C','Billed Cubic',
'G','Billed Gallons',
'A','Billed OMR',
'R','Collected',
'U','Total Unpaid Bills') mtr_type,
SUM(CASE
WHEN mtr_type NOT IN ('C','G','A','R','U') THEN 0 WHEN to_char(bill_month,'MMYYYY') = '012011' AND mtr_type = 'C' THEN cubic
WHEN mtr_type NOT IN ('C','G','A','R','U') THEN 0 WHEN to_char(bill_month,'MMYYYY') = '012011' AND mtr_type = 'G' THEN (gallons)
WHEN mtr_type NOT IN ('C','G','A','R','U') THEN 0 WHEN to_char(bill_month,'MMYYYY') = '012011' AND mtr_type = 'A' THEN amounts
WHEN mtr_type NOT IN ('C','G','A','R','U') THEN 0 WHEN to_char(bill_month,'MMYYYY') = '022011' AND mtr_type = 'R' THEN amounts
WHEN mtr_type NOT IN ('C','G','A','R','U') THEN 0 WHEN to_char(bill_month,'MMYYYY') <= '022011' AND mtr_type = 'U' THEN amounts ELSE 0 END) January
FROM tmp_paew_month_gafu1 WHERE ministry_code IN
(SELECT DISTINCT ministry_code FROM tmp_paew_month_gafu GROUP BY ministry_code)
GROUP BY ministry_code,mtr_type
ORDER BY 1,3 DESCSAMPLE OUTPUT:
MINISTRY_CODE MTR_TYPE JANUARY
001 Billed Gallons 5326252
001 Billed Cubic 24210.2363636364
001 Total Unpaid Bills 16402.5
001 Billed OMR 13287.36
001 Collected 10410.508
002 Total Unpaid Bills 12089.99
002 Billed Cubic 0
002 Billed OMR 0
002 Billed Gallons 0
002 Collected 0
003 Total Unpaid Bills 27418.711
2nd query:
SELECT ministry_code,
DECODE(mtr_type, 'C','Billed Cubic',
'G','Billed Gallons',
'A','Billed OMR',
'R','Collected',
'U','Total Unpaid Bills') mtr_type,
SUM(CASE
WHEN mtr_type NOT IN ('C','G','A','R','U') THEN 0 WHEN to_char(bill_month) = to_date('012011','MMYYYY') AND mtr_type = 'C' THEN cubic
WHEN mtr_type NOT IN ('C','G','A','R','U') THEN 0 WHEN to_date(bill_month) = to_date('012011','MMYYYY') AND mtr_type = 'G' THEN (gallons)
WHEN mtr_type NOT IN ('C','G','A','R','U') THEN 0 WHEN to_date(bill_month) = to_date('012011','MMYYYY') AND mtr_type = 'A' THEN amounts
WHEN mtr_type NOT IN ('C','G','A','R','U') THEN 0 WHEN to_date(bill_month) = to_date('022011','MMYYYY') AND mtr_type = 'R' THEN amounts
WHEN mtr_type NOT IN ('C','G','A','R','U') THEN 0 WHEN to_date(bill_month) <= to_date('022011','MMYYYY') AND mtr_type = 'U' THEN amounts ELSE 0 END) January
FROM tmp_paew_month_gafu1 WHERE ministry_code IN
(SELECT DISTINCT ministry_code FROM tmp_paew_month_gafu GROUP BY ministry_code)
GROUP BY ministry_code,mtr_type
ORDER BY 1,3 DESCSample OUTPUT:
MINISTRY_CODE MTR_TYPE JANUARY
001 Total Unpaid Bills 106062.002
001 Billed OMR 0
001 Billed Cubic 0
001 Billed Gallons 0
001 Collected 0
002 Total Unpaid Bills 42023.243
002 Collected 0
002 Billed Gallons 0
002 Billed Cubic 0
002 Billed OMR 0
003 Total Unpaid Bills 133260.435

Answer 1

What is the sense of to_date(bill_month) in the second query?
bill_month is already a date. to_date function will expect a character parameter and convert it in a date value.
So what will happen? First the bill_month will be implicitely converted to a character value.
Second this character value will be converted back to a date. You gave no format string. So both conversions will be made will your setting of nls_date_format.
I can only guess what you really want. Perhaps you want to trunc the bill_month with trunc(bill_month,'MM').
Please give us
- table create script
- insert statements for test data
- required output for these test data
- description of your requirements
Edited by: hm on 22.11.2012 23:16