Arithmatic operations on fields in tables

Question

Hello Members,
My query might seem pretty basic, but seems uphill to me, as i have no idea how to perform arithmatic operations between fields in tables.
I have two tables temp1 and temp2. each of these tables have 2 fields : account no. and balance
the data would resemble the following, in each of the tables:
6.28899273 0
6.28899274 0
6.28899275 0
6.28899289 625.12
6.28899292 2666.24
Can you please give me an idea how to go about doing the following:
1. Compare the tables temp1 and temp2 for their respective account numbers and balances
2. Copy the results in a new table temp3, which will have 2 columns :
account no ( common to temp1 and temp2 )
deviation: subtracted balance of temp1 and temp2
regards,
novice82

Answer 1

user8635888 wrote:
Will be great if someone can give me a headstart on the following:
I have two tables temp1 and temp2 ( tables belonging to 2 different databases ) with two feilds
account no., ( Common in both tables )
balance ( float data type ).
1. I want to compare the balance feilds in temp1 and temp2.
print out the no. of accounts and percentage of match and mismatch.
2. output the a/c nos. whose balances dont match into a seperate table..Hello, this might be one approach for you:
First, identify those accounts that are different, and INSERT them into the table
INSERT /*+ APPEND */ INTO target_table
SELECT t1.account_no, t1.balance, t2.balance
  FROM table1 t1
  JOIN table2@db_link t2 ON (t1.account_no = t2.account_no)
WHERE NVL(t1.balance, -99999999999) != NVL(t2.balance, -99999999999);This assumes that -99999999999 can never be a valid balance.
Then, you can cycle through this table and print out the mismatches:
set serveroutput on size 100000 (or set serveroutput on size unlimited)
BEGIN
  FOR cur_rec IN (SELECT account_no, (t1.balance / CASE t2.balance WHEN 0 THEN NULL END) * 100 percent_mismatch FROM target_table)
    LOOP
      DBMS_OUTPUT.PUT_LINE('Account No: ' || cur_rec.account_no || ', Percent Mismatch: ' || cur_rec.percent_mismatch);
    END LOOP;
END;
/If you have a lot of records this may not be a feasible approach, and you'd be better looking at UTL_FILE if you to capture the anomolies in a file.
And alternative would be to do the whole lot in SQL:
set pagesize 5000
set linesize 300
set colsep ','
spool /tmp/your_file.csv
SELECT t1.account_no, t1.balance, t2.balance, (t1.balance / CASE t2.balance WHEN 0 THEN NULL END) * 100 percent_mismatch
  FROM table1 t1
  JOIN table2@db_link t2 ON (t1.account_no = t2.account_no)
WHERE NVL(t1.balance, -99999999999) != NVL(t2.balance, -99999999999);
user8635888 wrote:another question, that I have is, how does sql handle the computation, if a value in a particular field is divided by 0 You'll see from the code that if the divisor is zero, I'm substiting NULL, which means that the percent_mismatch will be NULL.