Group by platform AND hardware details (cpu architecture)

Unity Services
,
1

Hello,

In the Analytics dashboard, you can group by platform, which is great, but, can I group also by some kind of hardware specific?

Imagine I’d like to group by memory, or by cpu

More specific, what I want to check is how many users are using a device with Android with a CPU x86_64 and how many users are using a device with Android with a CPU ARM64

Thank you

2

Hi @Maeslezo

Yes, you can use specific event parameters, like CPU, memory etc. to group Analytics results.

There are a couple of ways to achieve this, the simplest is in the Data Explorer tool.

  1. Navigate to the Analytics > Data Explorer dashboard page.
  2. Create a New Report, choosing the “New Experience” option
  3. Click the “Add Event” option in the Measures Panel and select the name of the event that the parameter you are interested in is assigned to. In this case, select the clientDevice event, it contains the device parameters you are interested in.
  4. Change the “Aggregated By” selection to “Unique Users” to count the number of user devices rather than the number of events.
  5. Click on the “Dimensions” dropdown list in the Dimensions panel, scroll to the bottom and select the Event Parameter option.
  6. Select the parameter you are interested in from the newly displayed “Event Parameter” dropdown, in this case, I have selected cpuType
  7. Next, change the chart type to “Bar” and adjust date range filters as appropriate.
  8. Finally, Save your report and give it a name, so you can come back to it again at a later date, or add it to a Custom Dashboard.
    image
    image1263×756 58.1 KB

The second method uses the SQL Data Explorer. It is a bit more complex but is much more powerful.

  1. Navigate to the Analytics > SQL Data Explorer dashboard page
  2. Replace the SQL Query with one that will report on the specific parameter you are interested in, being careful to filter for the correct event name and a suitable date range to optimise the performance of your query.
  3. Click the Run button
  4. Scroll to the bottom of the page to view your results
  5. Use the charting options to create a visualization of your data. In this case, I chose cpuType on the X Axis and Unique Users on the Y Axis then selected a Column chart before Applying the charting selection.
  6. Finally, Save your query so you can return to it later or add it to a custom dashboard.
    image
    image921×785 39.7 KB

The SQL Query I entered was

select EVENT_JSON:cpuType::STRING as cpuType
, count(distinct USER_ID) as UniqueUsers
from EVENTS
where EVENT_DATE >= CURRENT_DATE-30
and EVENT_NAME = 'clientDevice'
group by cpuType
order by cpuType

The SQL Data Explorer lets you perform deeper analysis and reporting, check out the SQL Data Explorer documentation and the SQL Cookbook for some more advanced examples .

3 Likes
3

@Laurie-Unity Amazing explanation. Very detailed. Thanks a lot

2 Likes