Unlock a world of possibilities! Login now and discover the exclusive benefits awaiting you.
How to display the last transaction that has multiple rows. Example here is my data
transaction_id | transaction_date | account_name | items | amount |
9850840 | 5/10/2004 | Mikey | A | 10 |
9850840 | 5/10/2004 | Mikey | B | 15 |
9850840 | 5/10/2004 | Mikey | C | 20 |
4536281 | 5/11/2004 | Mikey | D | 10 |
4536281 | 5/11/2004 | Mikey | E | 15 |
4536281 | 5/11/2004 | Mikey | F | 20 |
7653428 | 5/12/2004 |
Mikey |
G | 10 |
7653428 | 5/12/2004 | Mikey | H | 15 |
7653428 | 5/12/2004 | Mikey | I | 20 |
7894566 | 5/11/2004 | James | A | 15 |
7894566 | 5/11/2004 | James | C | 20 |
7894566 | 5/12/2004 | James | D | 10 |
8664523 | 5/12/2004 | James | E | 20 |
8664523 | 5/12/2004 | James | E | 20 |
8664523 | 5/12/2004 | James | K | 30 |
I want to display in a chart table
account_name | items | amount |
Mikey | G | 10 |
Mikey | H | 15 |
Mikey | I | 20 |
James | E | 20 |
James | E | 20 |
James | K |
30 |
The Example I saw mostly sum up the amount, so it will only have 1 rec per account name, but i want to display the details.
master item last_transaction_id = FirstSortedValue(transaction_id,-transaction_date,account_name) will give me the last transaction_id per account name. How do I display all records associated with this transaction_id if I dont want to display the transaction_id. I can't get the set analysis syntax correctly, something like {<transaction_id = last_transaction_id>} account_name, {<transaction_id = last_transaction_id>} item, {<transaction_id = last_transaction_id>} Amount. i appreciate any help on this. Thanks.
First convert your transaction date from text to date/num format.
Date#(transaction_date,'MM/DD/YYYY') as transaction_date_Converted
Then Use this below expression with your dimension in table,
Sum({<transaction_date_Converted={"$(=Date(Max(transaction_date_Converted),'M/DD/YYYY'))"}>}amount)
It is not set analysis thing - it is the way Qlik objects are designed to work to always show you unique combinations of records + possibly aggregation.
The tricky bit indeed will be to have individual records as by default they will be grouped to dimensions used in the table object.
You say you dont want to use transaction_id in your table, but in reality you will have to retain granularity required to show all separate records.
Even on your example you dont have required granularity because your table
8664523 | 5/12/2004 | James | E | 20 |
8664523 | 5/12/2004 | James | E | 20 |
8664523 | 5/12/2004 | James | K | 30 |
in qlik UI object would be converted to this:
8664523 | 5/12/2004 | James | E | 40 |
8664523 | 5/12/2004 | James | K | 30 |
In order for you to retain granularity you would have to have unique identifier per each row you are wanting to display. Only then you could play around and possibly use calculated dimention with things like Dual() to show for example name but hold value of that unique identifier.
cheers