SUM rows by week syntax - sql

Question

I am using the following query
SELECT CONVERT(date,lot.Killdate) as KillDate
,lot.[LotNo]
,lot.[HotWeight]
,lot.[AverageYieldGrade]
,(lot.HeadShorn + lot.HeadUnshorn) as 'Total'
,(lot.HotWeight) / (lot.HeadShorn + lot.HeadUnshorn) as 'AvgWeight'
,date.Date
,date.WeekOfYear
FROM [LambLot].[dbo].[LotHeader] lot, Master_Dim.dbo.DateDim date
WHERE CONVERT(date,lot.KillDate) = date.Date
and lot.KillLocation = 1
AND lot.HeadShorn > 0
AND lot.HeadUnshorn > 0
AND date.Year = 2016
group by lot.LotNo, lot.HeadShorn, lot.HeadUnshorn, lot.HotWeight lot.AverageYieldGrade, lot.KillDate, date.Date, date.WeekOfYear
order by date.WeekOfYear asc
Here is a copy of the output
KillDate LotNo HotWeight AverageYieldGrade Total AvgWeight Date WeekOfYear
1 2016-01-04 102 21603.5 2.28 348 62.0790229885057 2016-01-04 2
2 2016-01-04 103 2305.3 1.42 53 43.4962264150943 2016-01-04 2
3 2016-01-04 105 1159 0 17 68.1764705882353 2016-01-04 2
4 2016-01-04 108 1493.6 0 39 38.2974358974359 2016-01-04 2
5 2016-01-04 109 2982.8 0 80 37.285 2016-01-04 2
What i would like to do is sum each row into the weekofyear. Essentially giving me 52 rows of output, with sums of each value in each column shown. Is there a way to do this?

Answer 1

If you need to SUM by just WeekOfYear, you need to GROUP BY just week of year. You have to decide how you handle the other values, here's one way that may make sense (with some expected output it'd be easier to figure out what you're really looking for):
SELECT CONVERT(date, MAX(lot.Killdate)) as KillDate
,MAX(lot.[LotNo]) -- ?? Does this really belong in this query?
,SUM(lot.[HotWeight])
,AVG(lot.[AverageYieldGrade])
,(SUM(lot.HeadShorn) + SUM(lot.HeadUnshorn)) as 'Total'
,SUM(lot.HotWeight) / (SUM(lot.HeadShorn) + SUM(lot.HeadUnshorn)) as 'AvgWeight'
,MAX(date.Date)
,date.WeekOfYear
FROM [LambLot].[dbo].[LotHeader] lot, Master_Dim.dbo.DateDim date
WHERE CONVERT(date,lot.KillDate) = date.Date
and lot.KillLocation = 1
AND lot.HeadShorn > 0
AND lot.HeadUnshorn > 0
AND date.Year = 2016
group by date.WeekOfYear
order by date.WeekOfYear asc
Just remember: whatever columns you GROUP BY will be factored into how a distinct group is figured out. Anything you don't have in your GROUP BY but want in your output just needs some kind of aggregate function (such as MIN, MAX, AVG, SUM, COUNT, etc.).