Microsoft Fabric Community Conference 2025, March 31 - April 2, Las Vegas, Nevada. Use code MSCUST for a $150 discount.
Register nowThe Power BI DataViz World Championships are on! With four chances to enter, you could win a spot in the LIVE Grand Finale in Las Vegas. Show off your skills.
I have the following tables, and I need to find the average weight all the IDs sold in Table 2 based off the recorded weights stored in table 2. How do I do this?
Can you explain this a bit more with some data? I do not see any recorded weights in Table2. Why is there a bi-directional crossfilter between the two?
You could merge the two tables together into a new table in query editor. Then all the fields will be available in new table. I could be analyzing your situation incorrectly though.
Thanks for your reply.
This is a mocked up model of what I'm trying to accomplish, and unfortunately, in my actual use case, table 2 is too large to accomodate additional columns (file size is large enough and I'm trying to keep it from growing our of control).
Hi @Adam1125,
Since we can't see you actual situation. Could you please share us some sample data and its corresponding expected result if possible? Otherwise we don't know where to start.
And you want to find the average weight all the IDs sold in Table 2 based off the recorded weights stored in table 2. Then how could we know which IDs are sold? What's the logic to make this calculation?
Thanks,
Xi Jin.
Here's what some sample data of what the tables might look like, and what I'm trying to accomplish.
Sample Tables, with Average Weight of Items Sold (probably a measure I'd want to build as a helper)
Average Total Weight - the end result, with the ability to slice into individual locations as shown
User | Count |
---|---|
136 | |
70 | |
69 | |
54 | |
53 |
User | Count |
---|---|
207 | |
94 | |
64 | |
61 | |
57 |