Using table as filter in power bi

Hi all,

I am trying to solve this probably simple problem but came up with strange result. I have a tLevels table with different grouping levels like this:

Lev1 Lev2 ProductID

1 12 A

1 12 B

1 4321 C

2 133 D

2 12 E

then I have a sales tSales table:

ProductID Amount

A 100

A 10

B 200

B 150

C 80

D 300

E 260

The tables are related via ProductID. I was expecting to sum ProductIDs by Lev1 with

SumAmount = CALCULATE(SUM(tSales[Amount]),tLevels[Lev1])

However I get this strange result

Any ideas why is it not summing by Lev1 giving me the same result on all lines?

tjd
Impactful Individual
9 years ago
Anytime you see a result like this (all values returned are the same) it usually means that your table relationships are not set correctly between your tLevels and tSales tables. Try looking at the relationship flow between your tables and select the other choice (one or both).
zivhimmel
Resolver I
9 years ago
If you want to show Amount by Lev1,
the measure will simple be =SUM(Amount).
You don't even have to create a measure for it, just create a visual with tLevels[Lev1] and tSales[Amount]
and make sure the aggregation used for tSales[Amount] is SUM.

The problem with the measure you've created is that CALCULATE changes the context filter for the fields you are filtering by
(in this case tLevels[Lev1]). The second argument to CALCULATE is a flter. You just used tLevels[Lev1] as a filter,
which always returns TRUE, therefore all values are the same.

I hope that helps.

Ziv Himmelfarb
https://www.superfarb.com
- gvg
  Post Prodigy
  9 years ago
  Thanks for the comments. I actually tried to put up a table visual without any measures before posting this question. It returns even more strange result:
  
  My table relationship looks like:
  
  It doesn't help if Cross filter direction is Both or Single.
  - zivhimmel
    Resolver I
    9 years ago
    Any page level or report level filters in use?

Forum Discussion