Pivot table of summations from multiple tables [closed] - mysql

Question

Closed. This question needs details or clarity. It is not currently accepting answers.
Want to improve this question? Add details and clarify the problem by editing this post.
Closed 3 years ago.
Improve this question
I have problems extracting data from these tables. I have a statistics system where numbers are entered for employees every some days in the month - for the sake of example here are only two months but it will be a full year. I tried a join but I received doubled results + some employee can work on more accounts while others could work on only one...
employees:
+----+-------------+--+
| ID | Name | |
+----+-------------+--+
| 1 | John Doe | |
| 2 | Helen Price | |
| 3 | John Price | |
+----+-------------+--+
Account
+----+---------+--+
| ID | Name | |
+----+---------+--+
| 1 | fisher | |
| 2 | fisher1 | |
+----+---------+--+
January
+----+-------------+---------+----------+----------+----------+
| ID | Employee | Account | firstday | seconday | thirdday |
+----+-------------+---------+----------+----------+----------+
| 1 | John Doe | fisher | 2 | 4 | 6 |
| 2 | Helen Price | fisher | 1 | 1 | 1 |
| 3 | John Price | fisher1 | 1 | 2 | 1 |
| 4 | John Price | fisher | 1 | 1 | 1 |
+----+-------------+---------+----------+----------+----------+
February
+----+-------------+---------+-----------+----------+-----------+
| ID | Employee | Account | fourthday | eightday | secondday |
+----+-------------+---------+-----------+----------+-----------+
| 1 | John Doe | fisher | 1 | 4 | 6 |
| 2 | Helen Price | fisher | 1 | 1 | 2 |
| 3 | John Price | fisher1 | 1 | 2 | 1 |
| 4 | John Price | fisher | 1 | 1 | 1 |
+----+-------------+---------+-----------+----------+-----------+
My question is how can I write a query with the following result:
+----+-------------+---------+-----------------+------------------+--------+--+
| ID | Employee | Account | January(totals) | February(totals) | Totals | |
+----+-------------+---------+-----------------+------------------+--------+--+
| 1 | John Doe | fisher | 12 | 11 | 23 | |
| 2 | Helen Price | fisher | 3 | 4 | 7 | |
| 3 | John Price | fisher | 3 | 3 | 6 | |
+----+-------------+---------+-----------------+------------------+--------+--+

Answer 1

What you're wanting is a pivot-table of aggregate results from the different JOIN tables.
There are a couple of approaches, but one approach is using LEFT JOIN with aggregate subqueries to total your monthly results by each employee and account.
In your sample data, there is no correlation of Employee to Account outside of your Month tables. So the association needs to be created by using another subquery to create the correlation to join the monthly tables on. This will cause an association for every employee and every account unless a many-to-many table is created to create the association.
Lastly since your sample data is only showing a single Account (fisher), you would need to filter it somehow, which was not specified clearly in your question, so I added it in to the WHERE criteria.
Example: DB-Fiddle
SELECT
e.ID,
e.Name AS Employee,
e.Account,
COALESCE(janurary_totals.total, 0) AS `January(totals)`,
COALESCE(february_totals.total, 0) AS `February(totals)`,
COALESCE(janurary_totals.total, 0) + COALESCE(february_totals.total, 0) AS Totals
FROM (
SELECT e1.ID, e1.Name, a1.Name AS Account
FROM employees AS e1
LEFT JOIN Account AS a1
ON a1.ID IS NOT NULL
) AS e
LEFT JOIN (
SELECT
Employee,
Account,
SUM(firstday) + SUM(seconday) + SUM(thirdday) AS total
FROM January
GROUP BY Employee, Account
) AS janurary_totals
ON janurary_totals.Employee = e.Name
AND janurary_totals.Account = e.Account
LEFT JOIN (
SELECT
Employee,
Account,
SUM(fourthday) + SUM(eightday) + SUM(secondday) AS total
FROM February
GROUP BY Employee, Account
) AS february_totals
ON february_totals.Employee = e.Name
AND february_totals.Account = e.Account
WHERE e.Account = 'fisher'
Result
| ID | Employee | Account | January(totals) | February(totals) | Totals |
| --- | ----------- | ------- | --------------- | ---------------- | ------ |
| 1 | John Doe | fisher | 12 | 11 | 23 |
| 2 | Helen Price | fisher | 3 | 4 | 7 |
| 3 | John Price | fisher | 3 | 3 | 6 |
You could greatly simplify things, by refactoring the individual month tables into a single employee_entry table with an employee_id, account_id, amount and DATE columns, that you could use to query against to create the pivot-table. As it stands currently, there is no means of determining which year the values are recorded, which will cause confusion on subsequent year totals.
The single entry table would allow you to create a pivot-table using conditional aggregates.
SUM(CASE MONTH(date_column) WHEN # THEN amount_column ELSE 0 END) AS MonthTotal
Allowing you to filter and group the results as desired.
Example: DB-Fiddle
SELECT
e.`Name` AS Employee,
a.`Name` AS Account,
m.`Year`,
m.`January(totals)`,
m.`February(totals)`,
m.`Totals`
FROM (
SELECT
ee.employee_id,
ee.account_id,
YEAR(ee.`datetime`) AS `Year`,
SUM(CASE MONTH(ee.datetime) WHEN 1 THEN ee.amount ELSE 0 END) AS `January(totals)`,
SUM(CASE MONTH(ee.datetime) WHEN 2 THEN ee.amount ELSE 0 END) AS `February(totals)`,
SUM(ee.amount) AS `Totals`
FROM employee_entry AS ee
GROUP BY ee.employee_id, ee.account_id, `Year`
) AS m
LEFT JOIN employees AS e
ON e.ID = m.employee_id
LEFT JOIN Account AS a
ON a.ID = m.account_id
Result
| Employee | Account | Year | January(totals) | February(totals) | Totals |
| ----------- | ------- | ---- | --------------- | ---------------- | ------ |
| Helen Price | fisher | 2020 | 3 | 4 | 7 |
| John Doe | fisher | 2020 | 12 | 11 | 23 |
| John Price | fisher | 2020 | 3 | 3 | 6 |
| John Price | fisher1 | 2020 | 4 | 4 | 8 |
To perform the conversion, you create the employee_entry table, and insert your individual monthly tables into it, by specifying a date for each of the day columns.
CREATE TABLE employee_entry (
`id` INTEGER NOT NULL AUTO_INCREMENT,
`employee_id` INTEGER,
`account_id` INTEGER,
`amount` INTEGER,
`datetime` DATE NULL DEFAULT NULL,
PRIMARY KEY (`id`)
);
INSERT INTO employee_entry
(`employee_id`, `account_id`, `amount`, `datetime`)
(
SELECT e.ID, a.ID, j.firstday, '2020-01-01'
FROM `January` AS j
LEFT JOIN employees AS e
ON j.Employee = e.Name
LEFT JOIN Account AS a
ON j.Account = a.Name
)
UNION
(
SELECT e.ID, a.ID, j.seconday, '2020-01-02'
FROM `January` AS j
LEFT JOIN employees AS e
ON j.Employee = e.Name
LEFT JOIN Account AS a
ON j.Account = a.Name
)
UNION
(
SELECT e.ID, a.ID, j.thirdday, '2020-01-03'
FROM `January` AS j
LEFT JOIN employees AS e
ON j.Employee = e.Name
LEFT JOIN Account AS a
ON j.Account = a.Name
)
UNION
(
SELECT e.ID, a.ID, f.fourthday, '2020-02-04'
FROM `February` AS f
LEFT JOIN employees AS e
ON f.Employee = e.Name
LEFT JOIN Account AS a
ON f.Account = a.Name
)
UNION
(
SELECT e.ID, a.ID, f.eightday, '2020-02-08'
FROM `February` AS f
LEFT JOIN employees AS e
ON f.Employee = e.Name
LEFT JOIN Account AS a
ON f.Account = a.Name
)
UNION
(
SELECT e.ID, a.ID, f.secondday, '2020-02-02'
FROM `February` AS f
LEFT JOIN employees AS e
ON f.Employee = e.Name
LEFT JOIN Account AS a
ON f.Account = a.Name
);