SQL group by, let null value stay - sql

Question

I have an SQL table for transfer histories like this:
ID | Date_out | Date_in
---+----------+----------
1 | 01.01.15 |
2 | 01.01.15 | 13.05.15
2 | 16.07.14 | 01.01.15
4 | 28.07.15 |
4 | 16.07.14 | 28.07.15
You can say that item with ID 2 was given away on 16.07.14 and returned on 01.01.15 and after that it went out again on 01.01.15 and returned back on 13.05.15. So it's in my office.
The item with ID 4 was given away on 16.07.14 and returned on 28.07.15 after that it went out again on 28.07.15. So it's not in my office
Now I want a list with every Item which is in my office and a list with items which are not in my office. I have to use Microsoft Access.
result 1: Items in my office
ID | Date_out | Date_in
---+----------+----------
2 | 01.01.15 | 13.05.15
result 2: Items outside my office
ID | Date_out | Date_in
---+----------+----------
1 | 01.01.15 |
4 | 28.07.15 |
The problem is that I have the items multiple times. When I group by ID the empty dates disappear like and I still have multiple items:
SELECT
table.item_ID,
table.Date_out,
table.Date_in
FROM table
WHERE table.date_in<Now()
GROUP BY table.item_ID
Does anyone have an idea?

Answer 1

another solution for your problem
SELECT
id,
CASE
WHEN
SUM(CASE WHEN date_in IS NULL THEN 1 ELSE 0 END) > 0
THEN 'out'
ELSE 'in'
END
FROM
table_name
GROUP BY id;

Answer 2

DROP TABLE IF EXISTS my_table;
CREATE TABLE my_table
(ID INT NOT NULL
,Date_out DATE NOT NULL
,Date_in DATE NULL
,PRIMARY KEY(id,date_out)
);
INSERT INTO my_table VALUES
(1 ,'2015-01-01',NULL),
(2 ,'2015-01-01','2015-05-13'),
(2 ,'2014-07-16','2015-01-01'),
(4 ,'2015-07-28',NULL),
(4 ,'2014-07-16','2015-07-28');
SELECT x.id on_loan
FROM my_table x
JOIN
( SELECT id, MAX(date_out) max_date_out FROM my_table GROUP BY id) y
ON y.id = x.id
AND y.max_date_out = x.date_out
WHERE date_in IS NULL;
+---------+
| on_loan |
+---------+
| 1 |
| 4 |
+---------+
SELECT x.id in_stock
FROM my_table x
JOIN (SELECT id, MAX(date_out) max_date_out FROM my_table GROUP BY id) y
ON y.id = x.id
AND y.max_date_out = x.date_out
WHERE date_in IS NOT NULL;
+----------+
| in_stock |
+----------+
| 2 |
+----------+
or, less useful...
SELECT GROUP_CONCAT(CASE WHEN date_in IS NULL THEN x.id END) on_loan
, GROUP_CONCAT(CASE WHEN date_in IS NOT NULL THEN x.id END) in_stock
FROM my_table x
JOIN
( SELECT id, MAX(date_out) max_date_out FROM my_table GROUP BY id) y
ON y.id = x.id AND y.max_date_out = x.date_out;
+---------+----------+
| on_loan | in_stock |
+---------+----------+
| 1,4 | 2 |
+---------+----------+