Question

Table rows I’m using:

`factid` int(11) NOT NULL,
`segid` int(11) NOT NULL,
PRIMARY KEY (`factid`,`segid`)

---

Here’s the steps of what I’m trying to accomplish:

• I need the total number of unique segid values in the table.
• I need a single factid value.
• This factid value needs to have rows with many unique segid values.
• ‘Many’ being closest to half the above total unique segid values.

I hope this explains what I’m trying to accomplish. Help appreciated.

Answer 1

SELECT TOP 1
    facts.factid
  , facts.segids
  , segs.total_segids
FROM (
    SELECT factid
      , count(distinct segid) as segids
    FROM my_table
    GROUP BY factid
  ) facts,
  (
    SELECT count(distinct segid) as total_segids
    FROM my_table
  ) segs
ORDER BY abs(facts.segids - (segs.total_segids / 2))