Share feedback directly with Fabric product managers, participate in targeted research studies and influence the Fabric roadmap.
Sign up nowGet Fabric certified for FREE! Don't miss your chance! Learn more
Hi,
I have a new problem.
I am trying to visualize a question where a respondent could have multiple answers for multiple brands.
I have found solutions, but they seem to me very complicated and are actually not really practical.
Additionally I would like to summarize answers as a net count. which is also apparently not easily possible.
1.
Since I have to pivot all answers, and thus multiply the respondents by the number of possible answers, I have a problem with the base for percentage.
I solved this problem by making a base table and additionally a pivotized table with the multiplied respondents. Then I matched them via the model view. This seems to me to be very cumbersome. Is there another possibility?
2.
When I try to create overcodes, I always get the gross count of the data in the matrix ... which is logical. This is a pity, because I would like to use the functionality of drilling.
Now I have summarized the overcodes as an additional "answer" and so I get the correct values. Unfortunately I cannot use a drill anymore.
Is there a possibility to display the overcodes as a net count and still have the possibility to drill?
I want to have the ability to expand or close the Overcodes. But i dont want the sums to be a gross count.. but a net count.
like that:
Thank you 🙂
Solved! Go to Solution.
try this measure
Q N = CALCULATE(DISTINCTCOUNT(data_answers[ID_match]), data_answers[Value] = "quoted")Proud to be a Super User!
try this measure
Q N = CALCULATE(DISTINCTCOUNT(data_answers[ID_match]), data_answers[Value] = "quoted")Proud to be a Super User!
Oh yes thank you.. i think this is solving my overcode problem.
Do you have an idea for my base problem too?
I now have the base and the answers in two different tables. And this is getting quite messy when i have more of those data to visualize. Is there a way to do this more easily?
Thanks
i think you need to improve the data model.
in my opinion, you do not need to have both answer and answer_oc table.
Create a column in OC table
ITEM OC = SWITCH(data_answers_OC[Item],"OC_1-2","1 AND 2","oc_3-4-5","3 AND 4 AND 5",data_answers_OC[Item])Then you only need to create one group of measures.
Proud to be a Super User!
I now found out how to construct my base.
it is with the same trick. I just use distinctcount on my ID. So it only counts the respondend one time.
thanks 🙂
Share feedback directly with Fabric product managers, participate in targeted research studies and influence the Fabric roadmap.
| User | Count |
|---|---|
| 57 | |
| 52 | |
| 42 | |
| 18 | |
| 14 |
| User | Count |
|---|---|
| 111 | |
| 104 | |
| 35 | |
| 27 | |
| 27 |