Question: Pulling chunks of record by specifying percentage (%)

Question

Before I upload a sample of the data I was wondering if someone could answer a question for me so I know whether it is possible.
Let’s say I have a column with bunch of IDs and another column with their age.
Is there a query which allows you to define your data by %
Eg. I want to pull 200 IDs.
Out of those I want 30% aged 18 - 29; 40% aged 30 - 39; 30% aged 40 - 50
Possible?
Using:
Oracle Database 10g Enterprise Edition Release 10.2.0.5.0 - 64bi

Answer 1

Chloe_19 wrote:
IF I wanted to make it 50% male 50% female out of the 200, how and where would I add that in your code?Do you mean in each age group or just by gender? In each age group:
WITH t AS (
           SELECT  id,
                   age,
                   gender,
                   case
                     when age <= 29 then '18 - 29'
                     when age <= 39 then '30 - 39'
                     else '40 - 50'
                   end age_group,
                   row_number() over(
                                     partition by case
                                                    when age <= 29 then '18 - 29'
                                                    when age <= 39 then '30 - 39'
                                                    else '40 - 50'
                                                  end,
                                                  gender
                                     order by id
                                    ) rn
             FROM  table_name
             WHERE age between 18 and 50
SELECT  id,
        age
  FROM  t
  WHERE (age_group = '18 - 29' AND gender = 'M' AND rn <= 200 * 0.3 / 2)
     OR (age_group = '18 - 29' AND gender = 'F' AND rn <= 200 * 0.3 / 2)
     OR (age_group = '30 - 39' AND gender = 'M' AND rn <= 200 * 0.4 / 2)
     OR (age_group = '30 - 39' AND gender = 'F' AND rn <= 200 * 0.4 / 2)
     OR (age_group = '40 - 50' AND gender = 'M' AND rn <= 200 * 0.3 / 2)
     OR (age_group = '40 - 50' AND gender = 'F' AND rn <= 200 * 0.3 / 2)
/Just by gender:
WITH t AS (
           SELECT  id,
                   age,
                   gender,
                   row_number() over(
                                     partition gender
                                     order by id
                                    ) rn
             FROM  table_name
             WHERE age between 18 and 50
SELECT  id,
        age
  FROM  t
  WHERE (gender = 'M' AND rn <= 200 / 2)
     OR (gender = 'F' AND rn <= 200 / 2)
/SY.