MS Access Small Equivalent

Question

I have this working in Excel however it really needs moved into Access as that's where the rest of the database resides.

Its simply one table that contains Unique_ID, Seller and Fruit...

1    Chris      Orange
2    Chris      Apple
3    Chris      Apple
4    Sarah      Kiwi
5    Chris      Pear
6    Sarah      Orange

The end results should be displayed by Seller and then a list of each fruit sold (in the following example Robert has not sold any fruit, I do have a list of all sellers name however this could be ignored in this example as that I believe that will be easy to integrate.) They will only sell a maximum of 20 fruit.

Seller   1st          2nd        3rd       4th
Chris    Orange  Apple    Apple   Pear
Sarah   Kiwi       Orange
Robert

At the moment Excel uses Index, Match and Small to return results. Small is simply used on the Unique_ID to find the 1st, 2nd, 3rd, ect...smallest entries and is matched to each sellers name to build the above results.

As Access doesn't have a Small function I am at a loss! In reality there are over 100,000 records (minimum) with over 4000 sellers....they are also not fruit :)

Answer 1

To understand DCount better, use it is a SELECT query instead of a crosstab:

SELECT Sales.ID, Sales.Seller, Sales.Fruit, DCount([id],"sales","seller='" & [seller] & "' and id<=" & [id]) AS N
FROM Sales;

On each row, the last column is the DCount result. The syntax is DCount (field, source, expression) so what it does is count the IDs (field) in the Sales table (source) that match the expression - in other words, has the same seller as that row's record and an ID <= the current row's ID. So for Chris's sales, it numbers them 1 through 4, even though Sarah had a sale in the middle.

From this result, it's easy to take a Crosstab query that makes a table with seller in the row and N in the column - putting the sales in order for each seller the way you wanted to see them. The "First" function finds the first fruit for the combination of seller and N for each row and column of the result. You could just as easily use "Max" or "Min" here - any text function. Of course, there is only one record matching the seller row and the N column, but Crosstab queries require a function to evaluate and cannot use "Group by" for the field selected as a Value.

My 1st answer combines these steps - the select and the crosstab queries - in one query.

Hope this helps.

Answer 2

TRANSFORM First(Sales.Fruit) AS FirstOfFruit
SELECT Sales.Seller
FROM Sales
GROUP BY Sales.Seller
PIVOT DCount([id],"sales","seller='" & [seller] & "' and id<=" & [id]);

Where the table name is "Sales" and the columns are "ID", "Seller" and "Fruit"