Forum Discussion
Group before calculating the Quartiles
Hello,
I am looking to group my data by one of the columns and then calculate the IQR of a corresponding column. The goal will be to find Q1 and Q3 values for each group then use this to filter outliers in the dataset.
To return the q1 for each group I tried
Q1 = SUMMARIZE(Table1,[groupedColumn],"q1", PERCENTILE.INC(Table1[valueColumn], 0.25))
and recieved the error:
'The expression refers to multiple columns. Multiple columns cannot be converted to a scalar value.'
The end goal will be to Filter all values < Q1-(IQR * 1.5) and > Q3 + (IQR * 1.5)
Any help would be greatly appreciated. Thanks in advance!
Hey,
I'm sorry for the delay, and I'm also sorry for mentioning it would be become a little more complex ... At least not now.
I created a little pbix file, and I used the same the formula like you do, except a minor issue. Here is my formula
summarize to percentile = SUMMARIZE('samplePercentile','samplePercentile'[GroupColumn],"p", PERCENTILE.INC('samplePercentile'[Value],0.25) )And all works well, I also tried the formula on a real-life dataset, the same result, working well ...
I'm wondering if your [groupedcolumn] is a measure or column in Table1
So, I guess you have to share your data, being able to reproduce your issue (at least trying).
By the way, what version of Power BI Desktop are you using.
I'm using the latest July update: 2.48.4792.721
Regards
6 Replies
- TomMartens
Super User
Hey,
can you please provide some sample data how your data in the table Table1 looks like?
Cheers
Some sample data can be seen here, with the groups in the column on the left and the values on the right
Thanks!
- TomMartens
Ah,
I see ...
This will become a little more complex, currently I'm in a hurry. So it will take ~ 1.5 I will post again ;-)
