Ok sorry for the weird question title, I've trouble on putting words on what I am trying to accomplish.
I have an OrderLine Fact table. Each row has a product (Category/SubCategory/Product). I need to have a measure that returns the number of Orders (not OrderLines) which have one or more Orderlines in a ProductCategory. Since the Order Fact Table is not linked to Product, I have to get this data from the OrderLine Fact Table.
My knowledge of MDX is rather limited and from what I could think of, all I have found is to put the "Number of Orders" (1 or -1) measure for each OrderLine and have a degenerated dimension of OrderNumber in the OrderLine table. With this, I thought I could get the SUM of MIN of "Number of Orders" by OrderNumber in the OrderLine Fact Table. And it works :
WITH MEMBER [Measures].[Test] AS
SUM([Order Number].[Order Hierarchy].CURRENTMEMBER, MIN([Order Number].[Order Hierarchy].CURRENTMEMBER.CHILDREN, [Measures].[Number of Orders]))
SELECT
{[Measures].[Number of Orders],[Measures].[Test]} ON 0,
{[Order Number].[Order Number].&[110100000030-BLA-01745892],[Order Number].[Order Number].&[110100000031-BLA-0]} ON 1
FROM SalesCube
For an order with 3 lines, it will return to me 3 for [Measures].[Number of Orders] and 1 for [Measures].[Test].
My problem now is if I want to execute this query by Product Categories :
WITH MEMBER [Measures].[Test] AS
SUM([Order Number].[Order Hierarchy].CURRENTMEMBER, MIN([Order Number].[Order Hierarchy].CURRENTMEMBER.CHILDREN, [Measures].[Number of Orders]))
SELECT
{[Measures].[Number of Orders],[Measures].[Test]} ON 0,
{[Products].[Category Hierarchy].MEMBERS} ON 1
FROM SalesCube
it gives me the MIN of all OrderLines which is (-1) and sum it to -1. What I need is for the query to return the sum of min "Number of Orders" for each "Order Number" that contains the product category without having to use the "Order Number" on the row axis.
Does that make any sense?
As far as I understand, the best way to model this would be to have an "Order ID" in the order line fact table, and just define a distinct count on it as the measure [Measures].[Order Count]. There should be no need for complex MDX then.
If you need to subtract the orders with -1, maybe you could convert the -1 and 1 to an attribute with user readable values like "placed" and "reversed". Then the "Order Count" measure would get invisible, and a calculated measure be shown to the user which would be implemented as follows:
([Measures].[Order Count], [Order].[Order Status].[placed]) -
([Measures].[Order Count], [Order].[Order Status].[reversed])
Related
If anyone can help me figure out where I am going wrong with this SQL that would be great. Please see my attempt to answer it below. I have answer how I think it should be answered but I am very confused by the exam advice below, which says I should use a SUM function? I have googled this and I do not see how a SUM function can help here when I need get the product of two values in this case. Or am I missing something major?
Question: TotalValue is a column in Order relation that contains derived data representing total value (amount) of each order. Write a SQL SELECT statement that computes a value for this column.
My answer:
SELECT Product.ProductPrice * OrderLine.QuantityOrdered AS Total_Value
FROM Product,
OrderLine
GROUP BY Product;
Advice from exam paper:
This is a straightforward question. Tip: you need to use the SUM function. Also, note that you can take the sum of various records set using the GROUP BY clause.
Ok your question became a lot clearer once I clicked on the the hyperlink (blue text).
Each order is going to be made up of a quantity of 1 or more products.
So there could be 3 Product A and 5 Product B etc.
So you have to get the total for each product which is your Price * Quantity, but then you need to add them all together which is where the SUM comes in.
Example:
3 * ProductA Price (e.g. €5) = 15
5 * ProductB Price (e.g. €4) = 20
Total Value = 35
So you need to use the Product, Order and OrderLine tables.
Something like (I haven't tested it):
SELECT SUM(Product.ProductPrice * OrderLine.QuantityOrdered) FROM Product, Order, OrderLine
WHERE Order.OrderID = OrderLine.OrderID
AND Product.ProductID = OrerLine.ProductID
GROUP BY Order.OrderID
This should return rows containing the totalValue for each order - the GROUP BY clause causes the SUM to SUM over each group - not the entire rows.
For a single order you would need add (before the GROUP BY) "AND Order.OrderID = XXXXX" where XXXXX is the actual orders OrderId.
I don't understand why my sql is not running,
it pop out a window say
"Your query does not include the specified expression ' SUM(SaleRecord.Number)*(product.Price' as part of an aggregate function"
SELECT SUM(SaleRecord.Number)*(Product.Price) AS TotalIncome
FROM Product, SaleRecord
WHERE Product.ProductID=SaleRecord.SaleProduct;
Product.Price is not part of the aggregate. Presumably, you intend:
SELECT SUM(SaleRecord.Number * Product.Price) AS TotalIncome
FROM Product INNER JOIN
SaleRecord
ON Product.ProductID=SaleRecord.SaleProduct;
Note that I also fixed the archaic join syntax.
You asked in my previous answer:
"thank you, I just make some mistake, now it is working. And sorry to
bother you more, I want to select the product who sell the most out,
how can I do it, I try to add MAX(xxx) on it, and it don't work"
Now, I am by no means an expert, but there are two processes going on. Your language is confusing so I'm going to assume you want to know which product sells the most in $$ terms (rather than count. For example, you might sell 1,000 $0.50 products, equallying $500 total sales, or 10 $500 products, totallying $5000. If you want the count or the dollar value, then the method changes slightly).
So the first process is to get the total sales of each product, which I outlined above. Then you want to nest that inside a second query, where you then select the max. I'll give you the code and then explain it:
SELECT ProductID, MAX(TotalSale)
FROM (
SELECT P.ProductID, SUM(S.Number)*P.Price AS TotalSale
FROM Products as P, SaleRecords as S
WHERE product.Productid = SaleRecord.SaleProduct
GROUP BY Product.ProductID
)
It's easiest to imagine this as querying a query. Your first query is in the FROM() statement. That will run and give you the output of total sale per product. Then the second query is ran (the top most SELECT line) that selects the productID and the sale amount that is the largest among all the products.
Your teacher may not like this since nesting queries is a little advanced (though completely intuitive IMO). Hopefully this helps!
You brackets are wrong - for each row you want to multiply the price by the number, and only then sum them:
SELECT SUM(SaleRecord.Number * Product.Price) AS TotalIncome
FROM Product, SaleRecord
WHERE Product.ProductID = SaleRecord.SaleProduct;
You have a bracket error:
SELECT SUM(SaleRecord.Number * Product.Price) AS TotalIncome
FROM Product INNER JOIN
SaleRecord ON Product.ProductID = SaleRecord.SaleProduct;
This is because you're not indicating which column to group by. The line you wrote is:
SUM(SaleRecord.Number) * Product.Price
Which sums all of the sale quantities (regardless of differences in product ID) and multiplies it by the price right? Well what if you have multiple products with different prices? Basically, you are doing a one to many match, where you have a total that is the sum of all the sales, multiplied by multiple prices. What you need is a group by command. I would modify your code to say:
SELECT Product.ProductID, SUM(SaleRecord.Number)*Product.Price AS TotalSales
FROM Product, SaleRecord
WHERE product.Productid = SaleRecord.SaleProduct
GROUP BY Product.ProductID
That should take care of it, telling the dbms to group each product together, sum the number of sales and then multiply by the price of that product.
You can nest that inside another query to get total Income:
SELECT SUM(TotalIncome)
FROM ( **the above code here)
EDIT: Or you can do it like the ways listed above where your query creates a TotalIncome for each ORDER, and then sums them all together. my way creates a total sale for each PRODUCT and then sums all the products
In my product dimension, I have an attribute called CustomerRating which is a string. Possible values are "1", "2", "3", and "4".
I want to turn this attribute into a Measure that averages ratings for collections of products.
with member [Measures].[Product Rating] as
( [Product].[Project Name].CurrentMember.Properties("CustomerRating"))
select [Measures].[Product Rating] on columns
from [MyCube]
This query produces an error - I suspect because I'm dealing with a string.
How do I turn customer rating into a measure that is an average rather than a sum?
The best approach would be to add an int column to the source table of the dimension, maybe just as a named calculation in the DSV. Then you would add a measure group on the dimension table, and define a measure rating_sum that sums this column, and a count measure in this measure group.
Then define a calculated measure as rating_sum / count.
If everything works, make the two measures rating_sum and count invisible.
Not tested but I'm wondering if this errors?
WITH
MEMBER [Measures].[Product Rating] AS
CInt([Product].[Project Name].CurrentMember.Properties("CustomerRating"))
SELECT
{[Measures].[Product Rating]} ON COLUMNS
From [MyCube]
Alright so say I have a 'product_catalog', and 'orders' tables. Each order has the product_catalog_id as a foreign key. What I want to return as the query results is the product_code (name of the product associated with a specific product_catalog_id) + a count of how many of each product_code have been ordered. That's easy enough with something like this (Oracle SQL):
SELECT pc.product_code,
COUNT(*) as count
FROM orders o
join product_catalog pc on pc.product_catalog_id = o.product_catalog_id
GROUP BY pc.product_code
ORDER BY count DESC;
but I also want to print various pieces of information from the order table such as total of all monthly charges for that product_code. That would seem easy enough with something like this:
(o.monthly_base_charge*count(*)) as "Monthly Fee"
but the problem is that there have been various monthly fees for the same product_code over time. If I add the above line in and add 'o.monthly_base_charge' to the group by statement, then it will print out a unique row for every variation of pricing for that product_code. How do I get it to ignore those price variations and just add together every entry with that product code?
It is a little unclear what you are asking. My best guess is that you want the sum of the monthly base charge:
SELECT pc.product_code,
COUNT(*) as count,
sum(o.monthly_base_charge) as "Monthly Fee"
FROM orders o join
product_catalog pc
on pc.product_catalog_id = o.product_catalog_id
GROUP BY pc.product_code
ORDER BY count DESC;
I'm not sure if this is exactly what you want. What happens if you have two orders in the same month for the same product?
You may need to do something like this since SQL will not be able to know which monthly base charge to multiply by the count.
SELECT pc.product_code,
COUNT(*) as count,
(min(o.monthly_base_charge)*count(*)) as "Monthly Fee"
FROM orders o
join product_catalog pc on pc.product_catalog_id = o.product_catalog_id
GROUP BY pc.product_code
ORDER BY count DESC;
Or you will need to add o.monthly_base_charge to the group by in order for sql to know how to determine the count()
GROUP BY pc.product_code, o.monthly_base_charge
SELECT
NON EMPTY {[Measures].[Fact Order Count]}
ON COLUMNS,
{ ([Front Manager].[Front Manager Id].[Front Manager Id].ALLMEMBERS * [Order Type].[Order Type].[Order Type].ALLMEMBERS ) }
ON ROWS
FROM
[TEST_DW] CELL PROPERTIES VALUE
So, I have three columns in the output:
Front Manager, Order Type, Order Count
The above query shows me the counts for each manager and order type combination. I need a fourth column which would be a percentage of the types of orders for each front manager.
So, if there are four types of orders (A, B, C, D), and a manager had 25 of each order totaling 100. The fourth column would read 25%.....
I have scoured the web on how to do this, but have really come up short on this one. Any direction on this would be greatly appreciated, I am definitely new to MDX. Thanks.
What you're looking for are MDX Calculated members.
Let's assume the member for order A is called : [Order Type].[Order Type].[Order A] and we want to calculate the percentage from the total.
WITH
MEMBER [Order A] AS ([Order Type].[Order Type].[Order A],[Measures].[Fact Order Count]) / ([Measures].[Fact Order Count]) , FORMAT_STRING = 'Percent'
SELECT
{[Measures].[Fact Order Count],[Measures].[Order A]} on 0
...
What is important in the calculated members is that you can evaluate any MDX tuple (e.g ([Order Type].[Order Type].[Order A],[Measures].[Fact Order Count]) ). This changing if needed the values coming from the pivot axis (defined in on 0 and on 1..). Note you can add calculated members for the measures as well as the other dimensions.