1. Accueil
  2. Question et réponse
  3. Obtenir la moyenne pondérée dans une catégorie avec plusieurs sous-catégories

Obtenir la moyenne pondérée dans une catégorie avec plusieurs sous-catégories

2017-07-09 5 views
0

Joe possède un magasin de location de DVD dans le monde entier. Il a des magasins dans le monde entier dans différentes régions. Chaque mois, il aime voir si son équipement dans ses magasins est à jour en trouvant le pourcentage de son inventaire avec un certain statut de "Actuel", "Besoins Mise à Jour" et "Doit Partir" pour chaque région. Cela permet à ses clients d'avoir la meilleure expérience dans ses magasins. Bill, son assistant, veut construire une requête SQL qui montrera à Joe le pourcentage d'équipement dans chaque région et leur statut. Le type avant lui a utilisé un excel pour faire cela mais Bill pense SQL et ensuite un rapport de SSRS qui permettrait à Joe de choisir quelle région il veut voir l'impressionnerait. Il veut faire passer dynamiquement de mois en trimestres à mesure que l'année avance dans le rapport. C'est ce que le projet de loi a été en mesure de trouver à ce jour:Obtenir la moyenne pondérée dans une catégorie avec plusieurs sous-catégories

with 
COUNTS as (
select distinct 
count(distinct cnts.[Asset ID]) [Equipment Counts] 
,cnts.[Month] 
,cnts.[Equipment Status] 
,'Q' + convert(varchar(20),datepart(qq,[Month])) [Quarter] 
,year(cnts.[Month]) [Year] 
,cnts.[Region] 
from TABLEIMPORT cnts 
group by cnts.[Month], cnts.[Equipment Status], cnts.[Region] 

), 

TOT as (
select distinct 
count(distinct tot.[Asset ID]) [Equipmnet Total] 
,tot.[Month] 
,'Q' + convert(varchar(20),datepart(qq,[Month])) [Quarter] 
,year(tot.[Month]) [Year] 
from TABLEIMPORT tot 
group by tot.[Month] 
), 

DATA as (
select 
c.[Month] 
,c.[Equipment Counts]/cast(t.[Equipment Total] as decimal (18,2)) [Percents] 
,c.[Equipment Status] 
,c.Quarter 
,c.Year 
,c.[Region] 
from counts c 
    inner join TOT t on t.[Month] = c.[Month] 
    ), 

QUARTERS as (
select distinct 
d.Quarter [Month] 
,avg(d.Percents) [Percents] 
,d.[Equipment Status] 
,d.Quarter 
,d.Year 
from data d 
group by d.[Equipment Status], d.Quarter, d.Year 
), 

OWN as (
select distinct 
o.[Region] 
,case 
    when o.[Region] in ('Northeast', 
           'Southeast', 
           'Midwest', 
           'Northwest', 
           'Southwest', 
           'Mexico', 
           'Canada',) 
     then 'North America' 
    when o.[Region] like 'Europe' 
     then 'Europe' 
    when o.[Region] like 'Africa' 
     then 'Africa' 
    when o.[Region] like 'Asia' 
     then 'Asia' 
    when o.[Region] like 'Australia' 
     then 'Australia' 
    else isnull(o.[Region], 'No Region') 
    end [SelectRegion] 

    from TABLEIMPORT o 

), 

QTRMNTH as (
select distinct 
q.Quarter [Month] 
,q.Percents [Percents] 
,d.[Equipment Status] 
,d.Quarter 
,d.Year 
,o.SelectRegion 
from data d 
    inner join OWN o on o.[Region] = d.[Region] 
    inner join QUARTERS q on q.Quarter = d.Quarter and q.year = d.Year and q.[Equipment Status] = d.[Equipment Status] 
where d.[Month] < DATEADD(qq, DATEDIFF(qq, 0, GETDATE()) - 1, -1) and d.[Month] > dateadd(MONTH,-12,GETDATE()) 


UNION 

select distinct 
convert(varchar(3),datename(month, d.[Month])) [Month] 
,d.Percents [Percents] 
,d.[Equipment Status] 
,d.Quarter 
,d.Year 
,o.SelectRegion 
from data d 
     inner join OWN o on o.[Region] = d.[Region] 

where d.[Month] > DATEADD(qq, DATEDIFF(qq, 0, GETDATE()) - 1, -1) and d.Year = year(getdate()) 
) 

select * 
from QTRMNTH 
--where QTRMNTH.SelectOrg in (@SelectOrg)

comparaison à Excel d'avant, tout est correct et son SSRS fonctionne parfaitement tous, sauf pour l'Amérique du Nord, qui a des pourcentages qui sont légèrement de. Il pense qu'une moyenne pondérée pour les sous-régions d'Amérique du Nord permettrait de corriger ce pourcentage pour que le pourcentage soit correct parce que certaines sous-régions sont plus grandes que d'autres.

Comment Bill procéderait-il pour calculer les moyennes pondérées de chacune des régions en Amérique du Nord?

EDIT:

Briser les différentes régions

with 
OWN as (
select distinct 
o.[Region] 
,case 
    when o.[Region] in ('Northeast', 'Southeast', 'Midwest', 'Southwest', 
          'Mexico', 'Canada',) 
    then 'North America' 
when o.[Region] like 'Europe' 
    then 'Europe' 
when o.[Region] like 'Africa' 
    then 'Africa' 
when o.[Region] like 'Asia' 
    then 'Asia' 
when o.[Region] like 'Australia' 
    then 'Australia' 
else isnull(o.[Region], 'No Region') 
end [SelectRegion] 

from TABLEIMPORT o 
), 

COUNTS as (
select distinct 
count(distinct cnts.[Asset ID]) [Equipment Counts] 
,cnts.[Month] 
,cnts.[Equipment Status] 
,'Q' + convert(varchar(20),datepart(qq,[Month])) [Quarter] 
,year(cnts.[Month]) [Year] 
,o.[Region] 
from TABLEIMPORT cnts 
    inner join OWN o on o.[Area] = cnts.[Region] 
group by cnts.[Month], cnts.[Equipment Status], o.[Region] 

), 

COUNTNE as (
select distinct 
count(distinct cnts.[Asset ID]) [Equipment Counts] 
,cnts.[Month] 
,cnts.[Equipment Status] 
,'Q' + convert(varchar(20),datepart(qq,[Month])) [Quarter] 
,year(cnts.[Month]) [Year] 
,o.[Region] 
from TABLEIMPORT cnts 
    inner join OWN o on o.[Region] = cnts.[Region] 
where o.[Region] like '%Northeast%' 
group by cnts.[Month], cnts.[Equipment Status], o.[Region] 
), 

COUNTSE as (
select distinct 
count(distinct cnts.[Asset ID]) [Equipment Counts] 
,cnts.[Month] 
,cnts.[Equipment Status] 
,'Q' + convert(varchar(20),datepart(qq,[Month])) [Quarter] 
,year(cnts.[Month]) [Year] 
,o.[Region] 
from TABLEIMPORT cnts 
    inner join OWN o on o.[Region] = cnts.[Region] 
where o.[Region] like '%Southeast%' 
group by cnts.[Month], cnts.[Equipment Status], o.[Region] 
), 

COUNTMW as (
select distinct 
count(distinct cnts.[Asset ID]) [Equipment Counts] 
,cnts.[Month] 
,cnts.[Equipment Status] 
,'Q' + convert(varchar(20),datepart(qq,[Month])) [Quarter] 
,year(cnts.[Month]) [Year] 
,o.[Region] 
from TABLEIMPORT cnts 
    inner join OWN o on o.[Region] = cnts.[Region] 
where o.[Region] like '%Midwest%' 
group by cnts.[Month], cnts.[Equipment Status], o.[Region] 
), 

COUNSW as (
select distinct 
count(distinct cnts.[Asset ID]) [Equipment Counts] 
,cnts.[Month] 
,cnts.[Equipment Status] 
,'Q' + convert(varchar(20),datepart(qq,[Month])) [Quarter] 
,year(cnts.[Month]) [Year] 
,o.[Region] 
from TABLEIMPORT cnts 
    inner join OWN o on o.[Region] = cnts.[Region] 
where o.[Region] like '%Southwest%' 
group by cnts.[Month], cnts.[Equipment Status], o.[Region] 
), 

COUNTNw as (
select distinct 
count(distinct cnts.[Asset ID]) [Equipment Counts] 
,cnts.[Month] 
,cnts.[Equipment Status] 
,'Q' + convert(varchar(20),datepart(qq,[Month])) [Quarter] 
,year(cnts.[Month]) [Year] 
,o.[Region] 
from TABLEIMPORT cnts 
    inner join OWN o on o.[Region] = cnts.[Region] 
where o.[Region] like '%Northwest%' 
group by cnts.[Month], cnts.[Equipment Status], o.[Region] 
), 

COUNTMX as (
select distinct 
count(distinct cnts.[Asset ID]) [Equipment Counts] 
,cnts.[Month] 
,cnts.[Equipment Status] 
,'Q' + convert(varchar(20),datepart(qq,[Month])) [Quarter] 
,year(cnts.[Month]) [Year] 
,o.[Region] 
from TABLEIMPORT cnts 
    inner join OWN o on o.[Region] = cnts.[Region] 
where o.[Region] like '%Mexico%' 
group by cnts.[Month], cnts.[Equipment Status], o.[Region] 
), 

COUNTCN as (
select distinct 
count(distinct cnts.[Asset ID]) [Equipment Counts] 
,cnts.[Month] 
,cnts.[Equipment Status] 
,'Q' + convert(varchar(20),datepart(qq,[Month])) [Quarter] 
,year(cnts.[Month]) [Year] 
,o.[Region] 
from TABLEIMPORT cnts 
    inner join OWN o on o.[Region] = cnts.[Region] 
where o.[Region] like '%Canada%' 
group by cnts.[Month], cnts.[Equipment Status], o.[Region] 
), 

COUNTEU as (
select distinct 
count(distinct cnts.[Asset ID]) [Equipment Counts] 
,cnts.[Month] 
,cnts.[Equipment Status] 
,'Q' + convert(varchar(20),datepart(qq,[Month])) [Quarter] 
,year(cnts.[Month]) [Year] 
,o.[Region] 
from TABLEIMPORT cnts 
    inner join OWN o on o.[Region] = cnts.[Region] 
where o.SelectRegion like 'Europe' 
group by cnts.[Month], cnts.[Equipment Status], o.[Region] 
), 

COUNTAS as (
select distinct 
count(distinct cnts.[Asset ID]) [Equipment Counts] 
,cnts.[Month] 
,cnts.[Equipment Status] 
,'Q' + convert(varchar(20),datepart(qq,[Month])) [Quarter] 
,year(cnts.[Month]) [Year] 
,o.[Region] 
from TABLEIMPORT cnts 
    inner join OWN o on o.[Region] = cnts.[Region] 
where o.SelectRegion like 'Asia' 
group by cnts.[Month], cnts.[Equipment Status], o.[Region] 
), 

COUNTAF as (
select distinct 
count(distinct cnts.[Asset ID]) [Equipment Counts] 
,cnts.[Month] 
,cnts.[Equipment Status] 
,'Q' + convert(varchar(20),datepart(qq,[Month])) [Quarter] 
,year(cnts.[Month]) [Year] 
,o.[Region] 
from TABLEIMPORT cnts 
     inner join OWN o on o.[Region] = cnts.[Region] 
where o.SelectRegion like 'Africa' 
group by cnts.[Month], cnts.[Equipment Status], o.[Region] 
), 

COUNTaus as (
select distinct 
count(distinct cnts.[Asset ID]) [Equipment Counts] 
,cnts.[Month] 
,cnts.[Equipment Status] 
,'Q' + convert(varchar(20),datepart(qq,[Month])) [Quarter] 
,year(cnts.[Month]) [Year] 
,o.[Region] 
from TABLEIMPORT cnts 
    inner join OWN o on o.[Region] = cnts.[Region] 
where o.SelectRegion like 'Australia' 
group by cnts.[Month], cnts.[Equipment Status], o.[Region] 
), 

TOT as (
select distinct 
count(distinct tot.[Asset ID]) [Equipment Total] 
,tot.[Month] 
,'Q' + convert(varchar(20),datepart(qq,[Month])) [Quarter] 
,year(tot.[Month]) [Year] 
,o.[Region] 
from TABLEIMPORT tot 
    inner join OWN o on o.[Region] = tot.[Region] 
group by tot.[Month], o.[Region] 
), 

TOTNA as (
select distinct 
count(distinct tot.[Asset ID]) [Equipment Total] 
,tot.[Month] 
,'Q' + convert(varchar(20),datepart(qq,[Month])) [Quarter] 
,year(tot.[Month]) [Year] 
,o.[Region] 
from TABLEIMPORT tot 
    inner join OWN o on o.[Region] = tot.[Region] 
where o.SelectRegion like 'North America' 
group by tot.[Month], o.[Region] 
), 

TOTEU as (
select distinct 
count(distinct tot.[Asset ID]) [Equipment Total] 
,tot.[Month] 
,'Q' + convert(varchar(20),datepart(qq,[Month])) [Quarter] 
,year(tot.[Month]) [Year] 
,o.[Region] 
from TABLEIMPORT tot 
    inner join OWN o on o.[Region] = tot.[Region] 
where o.SelectRegion like 'Europe' 
group by tot.[Month], o.[Region] 
), 

TOTAS as (
select distinct 
count(distinct tot.[Asset ID]) [Equipment Total] 
,tot.[Month] 
,'Q' + convert(varchar(20),datepart(qq,[Month])) [Quarter] 
,year(tot.[Month]) [Year] 
,o.[Region] 
from TABLEIMPORT tot 
    inner join OWN o on o.[Region] = tot.[Region] 
where o.SelectRegion like 'Asia' 
group by tot.[Month], o.[Region] 
), 

TOTAF as (
select distinct 
count(distinct tot.[Asset ID]) [Equipment Total] 
,tot.[Month] 
,'Q' + convert(varchar(20),datepart(qq,[Month])) [Quarter] 
,year(tot.[Month]) [Year] 
,o.[Region] 
from TABLEIMPORT tot 
    inner join OWN o on o.[Region] = tot.[Region] 
where o.SelectRegion like 'Africa' 
group by tot.[Month], o.[Region] 
), 

TOTAUS as (
select distinct 
count(distinct tot.[Asset ID]) [Equipment Total] 
,tot.[Month] 
,'Q' + convert(varchar(20),datepart(qq,[Month])) [Quarter] 
,year(tot.[Month]) [Year] 
,o.[Region] 
from TABLEIMPORT tot 
    inner join OWN o on o.[Region] = tot.[Region] 
where o.SelectRegion like 'Australia' 
group by tot.[Month], o.[Region] 
), 

DATANE as (
select distinct 
cne.[Month] 
,cne.[Equipment Counts]/cast(T.[Equipment Total] as decimal (18,2)) [Percents] 
,cne.[Equipment Status] 
,cne.Quarter 
,cne.Year 
,cne.[Region] 
from COUNTNE cne 
    inner join TOTNA t on t.[Month] = cne.[Month] 
), 

DATASE as (
select distinct 
cse.[Month] 
,cse.[Equipment Counts]/cast(T.[Equipment Total] as decimal (18,2)) [Percents] 
,cse.[Equipment Status] 
,cse.Quarter 
,cse.Year 
,cse.[Region] 
from COUNTSE cse 
    inner join TOTNA t on t.[Month] = cse.[Month] 
), 

DATAMW as (
select distinct 
cmw.[Month] 
,cmw.[Equipment Counts]/cast(T.[Equipment Total] as decimal (18,2)) [Percents] 
,cmw.[Equipment Status] 
,cmw.Quarter 
,cmw.Year 
,cmw.[Region] 
from COUNTMW cmw 
    inner join TOTNA t on t.[Month] = cmw.[Month] 
), 

DATASW as (
select distinct 
csw.[Month] 
,csw.[Equipment Counts]/cast(T.[Equipment Total] as decimal (18,2)) [Percents] 
,csw.[Equipment Status] 
,csw.Quarter 
,csw.Year 
,csw.[Region] 
from COUNTSW csw 
    inner join TOTNA t on t.[Month] = csw.[Month] 
), 

DATANW as (
select distinct 
cnw.[Month] 
,cnw.[Equipment Counts]/cast(T.[Equipment Total] as decimal (18,2)) [Percents] 
,cnw.[Equipment Status] 
,cnw.Quarter 
,cnw.Year 
,cnw.[Region] 
from COUNTNW cnw 
    inner join TOTNA t on t.[Month] = cnw.[Month] 
), 

DATAMX as (
select distinct 
cmx.[Month] 
,cmx.[Equipment Counts]/cast(T.[Equipment Total] as decimal (18,2)) [Percents] 
,cmx.[Equipment Status] 
,cmx.Quarter 
,cmx.Year 
,cmx.[Region] 
from COUNTMX cmx 
    inner join TOTna t on t.[Month] = cmx.[Month] 
), 

DATACN as (
select distinct 
ccn.[Month] 
,ccn.[Equipment Counts]/cast(T.[Equipment Total] as decimal (18,2)) [Percents] 
,ccn.[Equipment Status] 
,ccn.Quarter 
,ccn.Year 
,ccn.[Region] 
from COUNTCN ccn 
    inner join TOTna t on t.[Month] = ccn.[Month] 
), 

DATAEU as (
select distinct 
ceu.[Month] 
,ceu.[Equipment Counts]/cast(t.[Equipment Total] as decimal (18,2)) [Percents] 
,ceu.[Equipment Status] 
,ceu.Quarter 
,ceu.Year 
,ceu.[Region] 
from COUNTEU ceu 
    inner join TOTEU t on t.[Month] = ceu.[Month] 
), 

DATAAS as (
select distinct 
cas.[Month] 
,cas.[Equipment Counts]/cast(t.[Equipment Total] as decimal (18,2)) [Percents] 
,cas.[Equipment Status] 
,cas.Quarter 
,cas.Year 
,cas.[Region] 
from COUNTAS cas 
    inner join TOTAS t on t.[Month] = cas.[Month] 
), 

DATAAF as (
select distinct 
caf.[Month] 
,caf.[Equipment Counts]/cast(t.[Equipment Total] as decimal (18,2)) [Percents] 
,caf.[Equipment Status] 
,caf.Quarter 
,caf.Year 
,caf.[Region] 
from COUNTAF caf 
    inner join TOTAF t on t.[Month] = caf.[Month] 
), 

DATAAUS as (
select distinct 
caus.[Month] 
,caus.[Equipment Counts]/cast(t.[Equipment Total] as decimal (18,2)) [Percents] 
,caus.[Equipment Status] 
,caus.Quarter 
,caus.Year 
,caus.[Region] 
from COUNTAUS caus 
     inner join TOTAUS t on t.[Month] = caus.[Month] 
), 

DATA as (
select * 
from DATANE dne 

UNION 

select * 
from DATAse dse 

UNION 

select * 
from DATAmw dmw 

UNION 

select * 
from DATAsw dsw 

UNION 

select * 
from DATAnw dnw 

UNION 

select * 
from DATAcn dcn 

UNION 

select * 
from DATAmx dmx 

UNION 

select * 
from DATAeu deu 

UNION 

select * 
from DATAas das 

UNION 

select * 
from DATAaf daf 

UNION 

select * 
from DATAaus daus 
), 

QUARTERS as (
select distinct 
d.Quarter [Month] 
,avg(d.Percents) [Percents] 
,d.[Equipment Status] 
,d.Quarter 
,d.Year 
from data d 
group by d.[Equipment Status], d.Quarter, d.Year 
), 

QTRMNTH as (
select distinct 
q.Quarter [Month] 
,q.Percents [Percents] 
,d.[Equipment Status] 
,d.Quarter 
,d.Year 
,o.SelectRegion 
from data d 
    inner join OWN o on o.[Region] = d.[Region] 
    inner join QUARTERS q on q.Quarter = d.Quarter and q.year = d.Year and q.[Equipment Status] = d.[Equipment Status] 
where d.[Month] < DATEADD(qq, DATEDIFF(qq, 0, GETDATE()) - 1, -1) and d.[Month] > dateadd(MONTH,-12,GETDATE()) 

UNION 

select distinct 
convert(varchar(3),datename(month, d.[Month])) [Month] 
,d.Percents [Percents] 
,d.[Equipment Status] 
,d.Quarter 
,d.Year 
,o.SelectRegion 
from data d 
    inner join OWN o on o.[Region] = d.[Region] 

where d.[Month] > DATEADD(qq, DATEDIFF(qq, 0, GETDATE()) - 1, -1) and d.Year = year(getdate()) 
) 

select * 
from QTRMNTH

Source

2017-07-09 programr

+1

Comment retravailleriez-vous ceci afin qu'il ne semble pas être devoirs? Qu'est-ce que Bill a essayé de faire pour résoudre le problème? Où est cette tentative de requête? –

+0

Ajout d'une tentative de publication – programr

+0

@programr Vous pouvez publier le schéma de vos tables? échantillon de données et résultats? –

Répondre

0

Cette requête ressemble un peu comme le petit déjeuner de chiens. Les principaux problèmes:

  1. SELECT DISTINCT ne sont pas nécessaires GROUP BY requêtes en tant que valeurs sont déjà distinctes. Je peux comprendre l'utilisation de DISTINCT dans COUNTS et TOT requêtes à condition que les données dans TABLEIMPORT ne sont pas normalisées.
  2. Total chance de commentaires expliquant l'intention de la requête.
  3. La dénomination des variables/alias est très mauvaise.

Je soupçonne que le problème avec la première requête est ici:

QTRMNTH as (
select distinct 
q.Quarter [Month] 
,q.Percents [Percents] 
,d.[Equipment Status] 
,d.Quarter 
,d.Year 
,o.SelectRegion 
from data d 
    inner join OWN o on o.[Region] = d.[Region] 
    inner join QUARTERS q on q.Quarter = d.Quarter and q.year = d.Year and q.[Equipment Status] = d.[Equipment Status] 
where d.[Month] < DATEADD(qq, DATEDIFF(qq, 0, GETDATE()) - 1, -1) and d.[Month] > dateadd(MONTH,-12,GETDATE())

Je suppose que le code ci-dessus est censé retourner des résultats trimestriels par « une plus grande région » [SelectRegion] basée sur des données mensuelles (data). Si tel est le cas, vous devez supprimer le DISTINCT et ajouter le GROUP BY.

Je pense qu'il serait également utile de supprimer DISTINCT de toutes les autres requêtes pour pouvoir repérer d'autres problèmes cachés.

Source

2017-07-09 10:35:19 Alex

 Questions connexes

  • Aucun problème connexe^_^