I need to create very long SQLite statements, pivots outputting thousands of columns.
I've resorted to doing this using string manipulation.
Just wondering if the SQLite API has a more elegant solution.
Example of desired output:
str = "SELECT sample_id
, SUM(CASE WHEN gear_id = 1 THEN counts ELSE 0 END) AS g1
, SUM(CASE WHEN gear_id = 2 THEN counts ELSE 0 END) AS g2
...
, SUM(CASE WHEN gear_id = N THEN counts ELSE 0 END) AS gN
FROM table1 GROUP BY sample_id"
Related
I have 3 tables.
Client,Documents and ClientDocuments.
The first one is the clients,where the information for each client are.
The second one is the documents, which document can go in the system.
The third one is the ClientDocuments, which client has which document.
Tables Here
I have to do a select where i get the information from the clients, and how many documents of the 3 types of the documents they have.
Example,Client 1 Have one document called 'Contrato Social',and 2 called 'Ata de Negociação'.
In the select must return every client and in the columns ContratoSocial returns 1,Ata Negociacao returns 2 and Aceite de Condições Gerais returns 0.
I did this to show.
select idFornecedor,
txtNomeResumido,
txtNomeCompleto,
--txtEmail,
--txtSenha,
bitHabilitado,
(SELECT Count(idDocumentosFornecedoresTitulo) FROM tbDocumentosFornecedores WHERE idDocumentosFornecedoresTitulo = 1) AS 'Contrato Social',
(SELECT Count(idDocumentosFornecedoresTitulo) FROM tbDocumentosFornecedores WHERE idDocumentosFornecedoresTitulo = 2) AS 'Ata de Negociação',
(SELECT Count(idDocumentosFornecedoresTitulo) FROM tbDocumentosFornecedores WHERE idDocumentosFornecedoresTitulo = 3) AS 'Aceite de Condições Gerais'
from dbo.tbFornecedores tbf
order by tbf.txtNomeResumido asc
returns this:
Returns of the query
But its just counting how many documents from that type is in the database, i want to filter for each client, how should i do?
Working answer:
select tbf.idFornecedor, tbf.txtNomeResumido, tbf.txtNomeCompleto,
tbf.bitHabilitado,
sum(case when idDocumentosFornecedoresTitulo = 1 then 1 else 0 end) as contrato_social,
sum(case when idDocumentosFornecedoresTitulo = 2 then 1 else 0 end) as Ata_de_Negociação,
sum(case when idDocumentosFornecedoresTitulo = 3 then 1 else 0 end) as Aceite_de_Condições_Gerais
from dbo.tbFornecedores tbf left join
tbDocumentosFornecedores df
on tbf.idFornecedor = df.idFornecedor
group by tbf.idFornecedor, tbf.txtNomeResumido, tbf.txtNomeCompleto, tbf.bitHabilitado
order by tbf.txtNomeResumido asc
You need some way of matching the rows in tbDocumentosFornecedores to the rows in tbFornecedores. Your question is not clear on what column is used for that, but I might guess something like idDocumentosFornecedore.
You could fix your query by using a correlation clause. However, I might instead suggest conditional aggregation:
select tbf.idFornecedor, tbf.txtNomeResumido, tbf.txtNomeCompleto,
tbf.bitHabilitado,
sum(case when idDocumentosFornecedoresTitulo = 1 then 1 else 0 end) as contrato_social,
sum(case when idDocumentosFornecedoresTitulo = 2 then 1 else 0 end) as Ata_de_Negociação,
sum(case when idDocumentosFornecedoresTitulo = 3 then 1 else 0 end) as Aceite_de_Condições_Gerais
from dbo.tbFornecedores tbf left join
tbDocumentosFornecedores df
on tbf.idDocumentosFornecedore = df.idDocumentosFornecedore -- this is a guess
group by tbf.idFornecedor, tbf.txtNomeResumido, tbf.txtNomeCompleto, tbf.bitHabilitado
order by tbf.txtNomeResumido asc
I have an access to Oracle server. There is a table on the Oracle server called Transactions which contains the following data:
I don't known the number of values, so we need to implement dynamic sql in Oracle.
I need to pivot that data so the results are:
Any suggestions?
You can use conditional aggregation:
select subno,
sum(case when offer = 'offer1' then 1 else 0 end) as offer1,
sum(case when offer = 'offer2' then 1 else 0 end) as offer2,
sum(case when offer = 'offer3' then 1 else 0 end) as offer3
from t
group by subno;
This question already has answers here:
How can I get multiple counts with one SQL query?
(12 answers)
Closed 4 years ago.
I am counting specific things in a SQL Server table. I am using multiple count queries, but I am wondering if it's possible to combine them into a single query with the column name and count numbers in a single table for display.
My queries are:
select count(*) as Ask_Count
from Pld_OpenSalesOrdersLines
where left(C_Item_ID, 1) = '*'
select count(*) as M_Count
from Pld_OpenSalesOrdersLines
where (left(C_Item_ID, 1) = 'M' and len(C_Item_ID) = 1)
select count(*) as MN_Count
from Pld_OpenSalesOrdersLines
where (left(C_Item_ID, 2) = 'MN' and len(C_Item_ID) = 2)
I tried a couple stupid things to combine them, but they were a failure. I honestly can't even begin to think how to combine them, maybe it's not possible?
Thanks!
You could use CASE expression to perform conditional aggregation:
select
COUNT(CASE WHEN LEFT(C_Item_ID,1)='*' THEN 1 END) AS Ask_Count,
COUNT(CASE WHEN LEFT(C_Item_ID,1)='M' AND LEN(C_Item_ID)=1 THEN 1 END) M_Count,
COUNT(CASE WHEN LEFT(C_Item_ID,2)='MN' AND LEN(C_Item_ID)=2 THEN 1 END) MN_Count
from Pld_OpenSalesOrdersLines
Use conditional aggregation:
select sum(case when LEFT(C_Item_ID,1) = '*' then 1 else 0 end) as count_1,
sum(case when LEFT(C_Item_ID,1) = 'M' AND LEN(C_Item_ID)=1 then 1 else 0 end) as count_2,
sum(case when LEFT(C_Item_ID,2) = 'MN' AND LEN(C_Item_ID)=2 then 1 else 0 end) as count_3
from Pld_OpenSalesOrdersLines;
I would write the logic like this, though:
select sum(case when C_Item_ID like '*%' then 1 else 0 end) as count_1,
sum(case when C_Item_ID = 'M' then 1 else 0 end) as count_2,
sum(case when C_Item_ID = 'MN' then 1 else 0 end) as count_3
from Pld_OpenSalesOrdersLines;
Doing a left() on a column and then checking the length is redundant. Just use =.
Table structures:
Solution_Votes:
ID int
SolutionID string
Vote int
Solution:
ID int
Solution
VotesUp
VotesDown
Code:
SELECT
*,
(SELECT SUM(CASE WHEN voteUp = 1 THEN 1 ELSE 0 END)
FROM Solutions_Votes) AS VoteCountUp,
(SELECT SUM(CASE WHEN voteDown = 0 THEN 1 ELSE 0 END)
FROM Solutions_Votes) AS VoteCountDown
FROM
Solution
When I run this query it gives me the count on each row for voteUpCount and voteDownCount. I need the count to be based on the solution ID so that each solution has its count of up votes and down votes. If anybody can help it would be appreciated. Thanks in advance!
Just use conditional aggregation. In your case this is simple:
select sv.solutionid,
sum(case when sv.voteUp = 1 then 1 else 0 end) as VoteCountUp,
sum(case when sv.voteDown = 0 then 1 else 0 end) as VoteCountDown
from solutions_votes sv
group by sv.solutionid;
You only need the solutions table if some solutions have no votes and you want to include them.
EDIT:
You would include solutions in various way. Here is one:
select s.*, ss.VoteCountUp, ss.VoteCountDown
from solutions s left join
(select sv.solutionid,
sum(case when sv.voteUp = 1 then 1 else 0 end) as VoteCountUp,
sum(case when sv.voteDown = 0 then 1 else 0 end) as VoteCountDown
from solutions_votes sv
group by sv.solutionid
) ss
on s.solutionid = ss.solutionid;
I am having trouble formulating a query to get the desired output.
This query involves one table and two columns.
First column bld_stat has 4 different values Private, public, Public-Abandoned, Private-Abandoned the other column bld_type, single_flr, multi_flr, trailer, Whs.
I need to get results that look like this:
So far I can get the first two columns but after that I have not been able to logically get a query to work
SELECT bld_stat, COUNT(grade) AS single_flr
FROM (SELECT bld_stat,bld_type
FROM bld_inventory WHERE bld_type = 'single_flr') AS grade
GROUP BY bld_stat,bld_type,grade
The term you are going for is pivoting. I think this should work...no need for the subquery, and I've changed your group by to only bld_stat
SELECT bld_stat,
sum(case when bld_type = 'singl_flr' then 1 else 0 end) AS single_flr,
sum(case when bld_type = 'multi_flr' then 1 else 0 end) AS multi_flr,
sum(case when bld_type = 'trailer' then 1 else 0 end) AS trailer,
sum(case when bld_type = 'whs' then 1 else 0 end) AS WHS
FROM bld_inventory
GROUP BY bld_stat