SSAS - Sum of Top or Bottom - sum

SELECT
[Measures].[Internet Sales Amount] ON COLUMNS
,Tail
(
[Date].[Calendar Year].[Calendar Year].MEMBERS
,2
) ON ROWS
FROM [Adventure Works];
Above MDX query gives me output as :
Year Internet Sales Amount
CY 2003 $9,791,060.30
CY 2004 $9,770,899.74
I understood how this query worked but I want to create Calculated measure in a cube which will always give me sum of bottom 2 years. How to do this? I am a newbie to SSAS. I am good at designing simple measures and dimensions but when it comes to using MDX, I am mostly stuck.
PS: I tried using TopCount, BottomCount etc but here I want to order by "Year", which is a dimension.
Any help would be appreciated.
Thanks,
Parry

The following query calculates a measure for sum the the last 2 years of the Date dimension:
WITH
MEMBER [Measures].[Sales from the last 2 Years]
AS Aggregate( Tail( [Date].[Calendar Year].[Calendar Year].Members, 2)
, [Measures].[Internet Sales Amount]
)
SELECT { [Measures].[Sales from the last 2 Years]
} ON COLUMNS
, { Tail( [Date].[Calendar Year].[Calendar Year].Members, 2)
} ON ROWS
FROM [Adventure Works]
Other interesting query, would be calculating a measure for the sum of each year and its previous:
WITH
MEMBER [Measures].[Sales from 2 years]
AS Aggregate( { [Date].[Calendar Year].CurrentMember.PrevMember
: [Date].[Calendar Year].CurrentMember }
, [Measures].[Internet Sales Amount]
)
SELECT { [Measures].[Internet Sales Amount]
, [Measures].[Sales from 2 years]
} ON COLUMNS
, NON EMPTY
{ [Date].[Calendar Year].[Calendar Year]
} ON ROWS
FROM [Adventure Works]
It does a sum, because the aggregation type of the measure [Measures].[Internet Sales Amount] is Sum, and the Aggregate function aggregates according to the measure's aggregation type.
Key Concepts in MDX
Aggregate function reference
MDX is a hard topic to grasp; if you're starting out, I recommend you read the book MDX Solutions, 2nd edition.

Related

MDX formatting results

I'm new to MDX querying. I am having trouble changing how the MDX output is formatted. I have made a similar example using the "Adventure Works Internet Sales Model". See below:
WITH
MEMBER [Measures].[Calculate YTD] AS
Sum
(
periodstodate([Date].[Calendar].[Year],[Date].[Calendar].CurrentMember)
,[Measures].[Internet Total Sales]
)
SELECT
{[Measures].[Internet Total Sales]
,[Measures].[Calculate YTD]
} ON COLUMNS,
[Date].[Calendar].[Month] ON ROWS
FROM [Adventure Works Internet Sales Model]
WHERE ([Date].[Date].&[2012-01-01T00:00:00]:[Date].[Date].&[2018-01-01T00:00:00])WHERE ([Date].[Date].&[2012-01-01T00:00:00]:[Date].[Date].&[2018-01-01T00:00:00])
the results looks like this:
MDX results
What I would like to see is that the "[Date].[Calendar].[Month]" row to be displayed as the end of month date (e.g. 31-Mar-2019)
Welcome to SO!
You could add in an additional column showing the last day of each month:
WITH
MEMBER [Measures].[Last Day of Month] AS
[Date].[Calendar].CurrentMember.LastChild.MEMBER_CAPTION
MEMBER [Measures].[Last Day of Month v2] AS
TAIL(EXISTING [Date].[Date].[Date].MEMBERS).ITEM(0).ITEM(0).MEMBER_CAPTION
MEMBER [Measures].[Calculate YTD] AS
Sum
(
periodstodate(
[Date].[Calendar].[Year],[Date].[Calendar].CurrentMember
)
,[Measures].[Internet Total Sales]
)
SELECT
{
[Measures].[Last Day of Month]
, [Measures].[Last Day of Month v2]
,[Measures].[Internet Total Sales]
,[Measures].[Calculate YTD]
} ON COLUMNS,
[Date].[Calendar].[Month]
ON ROWS
FROM [Adventure Works Internet Sales Model]
WHERE
(
[Date].[Date].&[2012-01-01T00:00:00]:
[Date].[Date].&[2018-01-01T00:00:00]
);
Output

I'm stuck with MDX query

i need sales amount of first month of each quarter
WITH SET [FIRSTMONTHOFQTR] AS
DESCENDANTS(
DESCENDANTS(
[Date].[Calendar].CURRENTMEMBER,
[Date].[Calendar].[Calendar Quarter]
),
[Date].[Calendar].[Month]
)
SELECT {
[Measures].[Sales Amount]
} ON COLUMNS,
{
[FIRSTMONTHOFQTR]
} ON ROWS
FROM [Adventure Works];
With above im getting each month but i need only first month. how can i filter that?
This is one approach:
WITH
SET [Qtrs] AS
[Date].[Calendar].[Calendar Quarter]
SET [FIRSTMONTHOFQTR] AS
Generate
(
[Qtrs] AS s
,Head(Descendants(s.CurrentMember,[Date].[Calendar].[Month]))
)
SELECT
{[Measures].[Sales Amount]} ON COLUMNS
,[FIRSTMONTHOFQTR] ON ROWS
FROM [Adventure Works];
It returns the following:

MDX query for bottomcount function

SELECT
[Measures].[Internet Sales Amount] ON COLUMNS
,BottomCount
(
NonEmpty([Customer].[Customer].[Customer].MEMBERS)
,10
,[Measures].[Internet Sales Amount]
) ON ROWS
FROM [Adventure Works]
WHERE
[Date].[Calendar].[Calendar Year].&[2005];
The above query is showing me result having last 10 customer names with NULL measure value. I am using Adventure works Cube.
As per my understanding "bottomcount" is not working fine with where clause in "mdx" query. I want to fetch last 10 customer names for 2005 with measure value and hopefully i am looking for solution without using keyword "DESCENDANTS".
Please let me know in case, I am doing something wrong.
Give this a try:
bottomcount(
nonempty(
[Customer].[Customer].[Customer].Members
,[Measures].[Internet Sales Amount]
)
,10
,[Measures].[Internet Sales Amount]
)
You were getting the Customer names with NULL because in OLAP you are doing the cartesian product between the 2 dimensions (Measures is a special dimension, but it still acts as a dimension) it does not matter that you don't have values it will still crossjoin the members in your 2 sets and unless you request just the nonempty ones it will still give you the NULLs.
And if you'd like their full internet sales amount then move some of the logic into the WITH clause:
WITH
SET bot AS
NonEmpty
(
[Customer].[Customer].[Customer].MEMBERS
,(
[Measures].[Internet Sales Amount]
,[Date].[Calendar].[Calendar Year].&[2005]
)
)
SELECT
[Measures].[Internet Sales Amount] ON COLUMNS
,BottomCount
(
bot
,10
,[Measures].[Internet Sales Amount]
) ON ROWS
FROM [Adventure Works];

Creating a measure by filtering out a set from an existing measure

I am trying to implement something as follows:
WITH MEMBER Measures.Test2 AS
Sum
(
{
[Date].[Calendar].[Calendar Year].&[2006]
,[Date].[Calendar].[Calendar Year].&[2007]
}
,[Measures].[Internet Sales Amount]
)
SELECT Measures.Test2 ON COLUMNS
FROM [Adventure Works];
But i want the new measure Test2 to be sliceable according to the Calendar Year dimension. So i want something like
SELECT {Measures.Test2} ON 0,
{[Date].[Calendar].[Calendar Year].[Calendar Year].MEMBERS} ON 1
FROM [Adventure Works];
This is giving the same value for both the years 2006 and 2007.
In essence i want to create a member by taking a subset of an existing measure and then using it for further calculations
This script is not valid mdx:
SELECT (Measures.Test2,[Date].[Calendar].[Calendar Year].[Calendar Year] ON COLUMNS
FROM [Adventure Works];
You have a single ( before Measures.
Also you look like you're about to add a tuple ON COLUMNS which is not allowed. Only sets are allowed on rows and columns:
SELECT
{Measures.Test2} ON 0,
{[Date].[Calendar].[Calendar Year].[Calendar Year].MEMBERS} ON 1
FROM [Adventure Works];
Try the following:
WITH
MEMBER Measures.Test2 AS
Sum
(
Intersect
(
{[Date].[Calendar].CurrentMember}
,{
[Date].[Calendar].[Calendar Year].&[2006]
,[Date].[Calendar].[Calendar Year].&[2007]
}
)
,[Measures].[Internet Sales Amount]
)
SELECT
[Measures].test2 ON COLUMNS
,{[Date].[Calendar].[Calendar Year].MEMBERS} ON ROWS
FROM [Adventure Works];
The above returns this:
Or maybe all you want is a subselect:
SELECT
[Measures].[Internet Sales Amount] ON COLUMNS
,{[Date].[Calendar].[Calendar Year].MEMBERS} ON ROWS
FROM
(
SELECT
{
[Date].[Calendar].[Calendar Year].&[2006]
,[Date].[Calendar].[Calendar Year].&[2007]
} ON 0
FROM [Adventure Works]
);

How to get an "(all)" value for a level in a user defined hierarchy in SSAS/MDX

I am trying to work out how I can get a total for a level in a user defined hierarchy. For example if I do (using Adventure works here):
SELECT [Measures].[Internet Sales Amount] ON 0,
([Date].[Calendar Year].ALLMEMBERS * [Date].[Month of Year].ALLMEMBERS) ON 1
FROM [Adventure Works]
I get totals for each pair of year/month values.
How can I rewrite this query using the Calendar user hierarchy in the Date dimensions? Eg something like:
SELECT [Measures].[Internet Sales Amount] ON 0,
([Date].[Calendar].[Year] * [Date].[Calendar].[Month]) ON 1
FROM [Adventure Works]
You can use
SELECT [Measures].[Internet Sales Amount] ON 0,
Hierarchize([Date].[Calendar].[Calendar Year] + [Date].[Calendar].[Month]) ON 1
FROM [Adventure Works]
This will show you each year containing its value (which is the sum of the months contained in it) before the months.
The + is the short form of Union in MDX, and Hierarchize sorts a set in hierarchical order, which means parents before children.
And if you want something similar to your first result, the closest thing I can imagine would be via calculated measures like this:
WITH Member Measures.Year AS
IIf([Date].[Calendar].CurrentMember.Level IS [Date].[Calendar].[Calendar Year],
[Date].[Calendar].CurrentMember.Name,
Ancestor([Date].[Calendar].CurrentMember, [Date].[Calendar].[Calendar Year]).Name
)
Member Measures.Month AS
IIf([Date].[Calendar].CurrentMember.Level IS [Date].[Calendar].[Month],
[Date].[Calendar].CurrentMember.Name,
'All Months'
)
SELECT {
Measures.Year,
Measures.Month,
[Measures].[Internet Sales Amount]
}
ON 0,
Hierarchize([Date].[Calendar].[Calendar Year] + [Date].[Calendar].[Month]) ON 1
FROM [Adventure Works]
Note that calculated members need not return numbers, they can as well return strings.
Using allmembers as you have it should return you the ALL line and then each year and each month. If you are interested in only the all line you can change you query to:
SELECT [Measures].[Internet Sales Amount] ON 0,
([Date].[Calendar Year] * [Date].[Month of Year]) ON 1
FROM [Adventure Works]
Notice how it is specifying only the Dimension and the hierarchy.
If you wanted to get the data without the all line use the following:
SELECT [Measures].[Internet Sales Amount] ON 0,
([Date].[Calendar Year].children * [Date].[Month of Year].children) ON 1
FROM [Adventure Works]