Distinct Select on Varchar2 Column

Question

Hi,
I have a unique problem - (no pun intended) - pls bare with me while I explain...
I have the following table (tbl_party) :
p_num VARCHAR2(30)
p_name VARCHAR2(100)
p_cert NUMBER
i_cat NUMBER
p_stat DATE
Test 1.
When I do a select count(distinct(p_num)) from tbl_party;
Return 6298 rows
Test 2.
When I do a select distinct(p_num) from tbl_party;
Return 6298 rows
Test 3.
When I do a select distinct(p_num), p_cert from tbl_party;
Return 6300 rows???
Can some1 pls enlighten me if there is a limitation on distinct/unique when using it on VARCHAR2??
Thanks in advance.
Jaco

Answer 1

I think in third test you get 6300 not 6298 because there are two additional value for the same values in p_num column.
See this:
SQL> ed
Wrote file afiedt.buf
  1  with t as
  2  (select 1 as col1, 'abc' from dual
  3  union all
  4  select 2, 'bcd' from dual
  5  union all
  6* select 1, 'bcd' from dual)
SQL> ed
Wrote file afiedt.buf
  1  with t as
  2  (select 1 as col1, 'abc' from dual
  3  union all
  4  select 2, 'bcd' from dual
  5  union all
  6  select 1, 'bcd' from dual)
  7* select distinct(col1) from t
SQL> /
      COL1
         1
         2
SQL> ed
Wrote file afiedt.buf
  1  with t as
  2  (select 1 as col1, 'abc' from dual
  3  union all
  4  select 2, 'bcd' from dual
  5  union all
  6  select 1, 'bcd' from dual)
  7* select count(distinct(col1)) from t
SQL> /
COUNT(DISTINCT(COL1))
                    2
SQL> ed
Wrote file afiedt.buf
  1  with t as
  2  (select 1 as col1, 'abc' as col2 from dual
  3  union all
  4  select 2, 'bcd' from dual
  5  union all
  6  select 1, 'bcd' from dual)
  7* select distinct(col1), col2 from t
SQL> /
      COL1 COL
         1 bcd
         1 abc
         2 bcdAs you can see there are two values in col2 for one value in col1. That's why you get 6300 not 6298 even if you use DISTINCT.
Peter D.