Reputation: 358
Hello and thanks for your help. I'm new to GQL but have good SQL experence and think I may be missing something small. I have 2 sheets i'm working with
Main sheet
Colum G
InstanceID
i-554532f4693fc6186
i-09554fcda5f2f3262
i-0047551ae514412d5
-
Data Sheet
Colum A Colum B
i-554532f4693fc6186 10.12
i-554532f4693fc6186 12.12
i-554532f4693fc6186 13.12
i-554532f4693fc6186 17.12
i-554532f4693fc6186 30.12
I am trying to write a query that will find all the rows that match the Instance ID in column G against the datasheet Column A and return the AVG of all the matches in column B, the top 5 max, and top 5 min.
I'm finding that I can't point the query to a cell for referencing the instance ID. Is there a way? I'm using this to try to get the max and it works for 1 but I ned the top 5 or any number.
=sort(query('HeC-Metrics'!A:B,"select max(B) Where A = 'i-044532f4693fc6186'"))
I'm OK needing to do different queries for each of the required results, AVG, min, max. I would also like to reference the cell in the G column so I don't have to manually enter the InstanceID.
Thanks your time. Stephen
Upvotes: 1
Views: 1937
Reputation: 34220
So it's just a case of getting the right syntax to use a cell value as a match in the query
=query(Sheet2!A:B,"select avg(B) where A='"&G2&"' group by A label avg(B) ''",1)
Note that you don't really need the group by if you already have a list of distinct ID's to compare against, but you can't have an aggregate like avg without it.
To get the bottom 5, you can use filter & sortn
=transpose(sortn(filter(Sheet2!B:B,Sheet2!A:A=G2),5))
(I have transposed the result to get it in a row (row 2) instead of a column)
or you could use a query
=transpose(query(Sheet2!A:B,"select B where A='"&G2&"' order by B limit 5 label B '' ",1))
Similarly to get the top 5 you could use
=transpose(sortn(filter(Sheet2!B:B,Sheet2!A:A=G2),5,,1,false))
or
=transpose(query(Sheet2!A:B,"select B where A='"&G2&"' order by B desc limit 5 label B '' ",1))
This begs the question of whether you could get these results (a) without needing a list of distinct values and (b) in a single array formula without copying down.
You could certainly get the distinct ID's and averages straight away from a query. Getting the top or bottom n values from a number of groups is much more difficult. I have attempted it in a previous question, but it requires a long and unwieldy formula.
Upvotes: 1