How to combine 2 result sets - sql

Question

To solve the below question, I want to combine 2 result sets (Query #1 + Query #2) using SQL.
I've tried UNION but it didn't work. Help!
Question:
Sample Data:
Query #1:
SELECT
COUNT(student_id),
MAX(registration_date),
MAX(lab1)
FROM grades;
Query #1 result:
Query #2:
SELECT
MAX(SUM(NVL(lab1, 0) + NVL(lab2, 0)))
FROM grades
GROUP BY student_id;
Query #2 result:

Answer 1

You had a few descent attempts. If you could edit your existing post and supply some sample data would help better show us what the underlying basis would be. Are there multiple rows per lab1 and/or lab2. Are there multiple classes represented? You want a sum, but doing a max?
You need a group by per student so each student shows of their own. So you had that. You also had max(), but also sum(). So hopefully this will be close to what you need.
SELECT
g.student_id,
max( g.registration_date ) NewestRegistrationDate,
max( coalesce( g.lab1, 0 )) HighestLab1,
max( coalesce( g.lab2, 0 )) HighestLab2,
max( coalesce( g.lab1, 0 ))
+ max( coalesce( g.lab2, 0 )) HighestBothLabs
FROM
grades g
group by
g.student_id
At least the above will give you the per-student basis as your grid image would represent, not the final answer which apparently wants the total student and the highest grade overall. That one, you WOULD want the count(), and ignore the group by as you want the answer over the entire set. But do you want the highest in each lab category? Or the highest per student with whatever the highest student's total labs are.
EX:
Student Lab1 Lab2 Total
1 75 92 167
2 74 100 174
Here, you have two students. The first has the highest lab1, but the second has highest total. You need to make that determination or at least get clarification from the teacher. If you want the highest by lab 1, you probably need something like wrapping the first query into the second. But that too may be over what you are learning. Using one result as basis for the outer such as.
select
count(*) TotalStudents,
Max( PerStudent.NewestRegistrationDate ) MostRecentRegistrationDate,
Max( PerStudent.HighestLab1 ) HighestFirstLab,
Max( PerStudent.HighestBothLabs ) HighestAllLabs
from
( SELECT
g.student_id,
max( g.registration_date ) NewestRegistrationDate,
max( coalesce( g.lab1, 0 )) HighestLab1,
max( coalesce( g.lab2, 0 )) HighestLab2,
max( coalesce( g.lab1, 0 ))
+ max( coalesce( g.lab2, 0 )) HighestBothLabs
FROM
grades g
group by
g.student_id ) PerStudent
The inner pre-query (alias PerStudent) keeps the data at the per-student level giving you the basis to get the highest per at the OUTER query level that does NOT have the group by. But at least running the first query to confirm the baseline of the data first. If you dont have that correct, the outer won't make sense either.

Answer 2

Thanks to DRapp, I solved it with the code below.
SELECT
COUNT(*) TotalStudents,
MAX(PerStudent.MostRecentRegistrationDate) MostRecentRegistrationDate,
MAX(PerStudent.HighestFirstLab) HighestFirstLab,
MAX(PerStudent.SumBothLabs) HighestBothLabs
FROM
(SELECT
MAX(g.registration_date) MostRecentRegistrationDate,
MAX(coalesce(g.lab1, 0)) HighestFirstLab,
SUM(coalesce(g.lab1, 0) + coalesce(g.lab2, 0)) SumBothLabs
FROM grades g
GROUP BY g.student_id) PerStudent;