https://www.gooddata.com/ logo
Join Slack
Powered by
So, a little more confusion around how to specify ...
# gd-beginners
m
So, a little more confusion around how to specify my MAQL calculations. I have a many to many relationship with Ethnicity for my sample model. When I have ethnicity added as only a filter, if I pick a single Ethnicity it calculates the population correctly, when I pick multiple as a filter, the overall population changes.
So the entire population is 15.2. Ignoring the multidimensional aspect. A response may have one or more ethnicities.
Bring in ethnicity on the display axis and we're still good.
However filtering by ethnicity drives up the population by the duplicate values.
I can ignore parent filters, for ethnicity, but that doesnt work for my population as we shockingly have more than two ethnicities and I still want to filter down.
I'm on NZ time so will be afk for a few hours 🙂
j
Hello Marcel, depending on your LDM and how these objects connect you can aggregate the visualization using The BY clause and its variations. Would you be able to share a screen shot of the metric used and the elements in the LDM?
m
image.png
image.png
I'll make it even simpler: Theres the below number of people in my survey group.
But when I filter by Ethnicity the number of people in my survey changes.
Because Good Data is counting the filtering as if the values were being displayed in the output.
I can ignore PARENT FILTERS, but I still want to filter the output by the two related ethnicities.
I've used the BY word here to "Ignore" Ethnicity. Works beautifully with the filter as well.
But take the dimension out of the rows, it roll on back up to the wrong number.
j
Hi Marcel, I believe this is covered in our documentation here: https://help.gooddata.com/doc/growth/en/data-integration/data-modeling-in-gooddata/many-to-many-in-logical-data-models/#double-count[…]n-an-mn-model You will need to edit the visualization accordingly in this type of LDM set up.
Open in Slack
PreviousNext
Powered by