Get average days open for issues per day (SQL) - sql

Question

I'm trying to create a chart in PowerBI that shows the average days open of issues in a database at each given time between two dates.
CREATE TABLE Issues
(IssueID int,IssueName varchar(10),created datetime, closed datetime);
INSERT INTO Issues
VALUES
(1,'a','2012-01-01 00:00:00', '2012-05-01 00:00:00'),
(2,'b','2012-03-01 00:00:00', '2012-06-01 00:00:00');
My first query shows all data in the database:
SELECT IssueID,
DATEDIFF(DAY,created,ISNULL(closed,GETDATE())) AS 'Days_Open'
FROM Issues
Results:
IssueID Days_Open
1 4
2 3
What I want to find is for each day, the issues open that day and average days open?
DECLARE #StartDate DATETIME
DECLARE #EndDate DATETIME
SET #StartDate = '2012-01-01'
SET #EndDate = '2012-07-01'
Ex:
Issues open each day
Date IssueID Days_Open
2012-01-01 1 0
2012-02-01 1 1
2012-03-01 1 2
2012-03-01 2 0
2012-04-01 1 3
2012-04-01 2 1
2012-05-01 1 4
2012-05-01 2 2
2012-06-01 2 3
Day 07 has no issues
Average
Date Average_Days_Open
2012-01-01 0 (1 issue just created)
2012-02-01 1 (1 issue - 1 day old)
2012-03-01 1 (2 issues - (2+0)/2 = 1)
2012-04-01 2 (2 issues - (3+1)/2 = 2)
2012-05-01 3 (2 issues - (4+2)/2 = 3)
2012-06-01 3 (1 issue - (0+3)/1 = 3)
2012-07-01 0 (Since there were no issues that day)
If i can get the data from both then I should be able to create a line chart in PowerBi similar to this:
Chart
Can someone please help out?

Answer 1

Based on syntax I guess it is SQL Server:
1) Query:
WITH CTE_DatesTable
AS
(
SELECT CAST('20120101' as datetime) AS [date]
UNION ALL
SELECT DATEADD(dd, 1, [date])
FROM CTE_DatesTable
WHERE DATEADD(dd, 1, [date]) <= '20120106'
)
SELECT [date] , i.IssueId, DATEDIFF(DAY,i.created,d.[date]) AS Days_Open
FROM CTE_DatesTable d
JOIN Issues i
ON d.date BETWEEN i.created AND i.closed
ORDER BY [date], IssueId
OPTION (MAXRECURSION 0);
DBFiddle Demo
2) Query (average):
WITH CTE_DatesTable
AS
(
SELECT CAST('20120101' as datetime) AS [date]
UNION ALL
SELECT DATEADD(dd, 1, [date])
FROM CTE_DatesTable
WHERE DATEADD(dd, 1, [date]) <= '20120106'
)
SELECT [date] ,AVG( DATEDIFF(DAY,i.created,d.[date]))
FROM CTE_DatesTable d
JOIN Issues i
ON d.date BETWEEN i.created AND i.closed
GROUP BY [date]
ORDER BY [date]
OPTION (MAXRECURSION 0);
DBFiddle Demo2
The whole idea is to generate calendar table.
If you need dates that are out of range you could use LEFT JOIN:
WITH CTE_DatesTable
AS
(
SELECT CAST('20120101' as datetime) AS [date]
UNION ALL
SELECT DATEADD(dd, 1, [date])
FROM CTE_DatesTable
WHERE DATEADD(dd, 1, [date]) <= '20120107'
)
SELECT [date] , COALESCE(AVG( DATEDIFF(DAY,i.created,d.[date])),0)
FROM CTE_DatesTable d
LEFT JOIN Issues i
ON d.date BETWEEN i.created AND i.closed
GROUP BY [date]
ORDER BY [date]
OPTION (MAXRECURSION 0);
DBFiddle Demo3