Forum Discussion
Calculating Weekly Bucket Difference
Hello Guys,
the following challenge:
For our inventory tracking, we build groups:
green = < 30 days
yellow = 30 - 90 days
red = > 90 days
In Power Bi we realized that with the SWITCH statement
Group by Age =
SWITCH(
True(),
fact_stock[age in days]>=90,"3. >=90 days",
fact_stock[age in days]>=30,"2. 30 - 90 days",
fact_stock[age in days]>=0,"1. <= 30 days" )
-> this statement work fine. the age in days is in database.
-> the fact_table is a weekly export with the following structure:
-> the unique ID is an Stock unit ID
date | stock unit id | value | age in days | group by age |
05.03.2023 | 12345679 | 100 | 23 | < 30 days |
05.03.2023 | 12345678 | 250 | 90 | > 90 days |
05.03.2023 | 12345677 | 100 | 89 | 30 - 90 days |
05.03.2023 | 12345676 | 50 | 33 | 30 - 90 days |
12.03.2023 | 12345679 | 50 | 30 | 30 - 90 days |
12.03.2023 | 12345678 | 50 | 97 | > 90 days |
12.03.2023 | 12345677 | null | null | consumed |
12.03.2023 | 12345676 | 25 | 40 | 30 - 90 days |
12.03.2023 | 12345684 | 75 | 2 | < 30 days |
Following questions i have to answer every monday:
- which stock units are new in comparison to previuos week
- which stock units are consumpted in comparison to previus week?
- which stock units changed the age bucket?
- for example: stock unit 12345678 changed from 30 - 90 days into bucket > 90 days
How should i build this in power bi?
2 Replies
To report on things that aren't there you need to use disconnected tables and crossjoins.
Status = SWITCH(TRUE(), min(Weeks[date])=CALCULATE(min(Weeks[date]),all(Weeks[date])),BLANK(), not isblank([val]) && isblank(calculate([Val],DATEADD(Weeks[date],-7,DAY))),"new this week", isblank([val]) && not isblank(calculate([Val],DATEADD(Weeks[date],-7,DAY))),"consumed", [Bucket]<>(calculate([Bucket],DATEADD(Weeks[date],-7,DAY))),"bucket change")see attached
Hi,
The last question is the most interesting one so i will take that up later. For the first 2 questions, could you kinldy share the following:
- Download link of the PowerBI file
- A Calendar table with a week column in there
- The expected result of the first 2 questions based on the sample data that you have shared in your post.
