If I have an event called "pageview X", how can I see the distribution of how often each property "id" occurs? I expect there are over 100k values for "id", so I'm trying to understand whether each ID occurs equally, or how many are more common (and by how much? twice as often? 10x?)
Hi Naomi J., Please stand by as I search the Mixpanel knowledge base for an AI-powered answer. I normally take about 15 seconds or so. In case AI cannot answer your question, feel free to reach out to Mixpanel support at mixpanel.com/get-support. If you have follow-up questions for this thread, please tag me again with your question: @kapa.ai <follow-up question> 👋
Feel free to ask a follow-up question once I've answered by mentioning me with @kapa.ai
To analyze the distribution of how often each property value for "id" occurs in your "pageview X" event, you can use Mixpanel's breakdown and aggregation features in the Insights report.
Here's how you can approach this:
1. Create an Insights Report: - Select the "pageview X" event as your metric. - For the measurement, use "Total Events" to count how many times each event occurred.
2. Breakdown by Property "id": - Add a breakdown by the "id" property. This will show you the count of "pageview X" events for each unique "id" value. - This breakdown will let you see which "id" values are most common and how their frequencies compare.
3. Analyze the Distribution: - With over 100k unique "id" values, the report will show you the distribution—some IDs may occur much more frequently than others. - You can use the distribution or percentiles aggregation options to understand how skewed the distribution is (e.g., are a few IDs much more common than the rest, and by how much).
4. Aggregation Options: - Mixpanel allows you to aggregate property values using functions like sum, average, median, percentiles, minimum, and maximum. For your use case, the "distribution" or "percentiles" options will help you see how many IDs are more common and by what factor (e.g., 2x, 10x more frequent) compared to others. This is especially useful for large cardinality properties like "id" Insights: Measurements.
Note: If you want to see the full distribution or export the data for further analysis (since the UI may limit the number of unique values displayed), you may need to use Mixpanel's API or JQL for more advanced queries.
This approach will help you determine whether each "id" occurs equally or if some are significantly more common than others, and by how much Events and Properties.
If you need to compute more advanced statistics or handle the large number of unique IDs programmatically, consider using JQL with groupBy on the "id" property to get the counts per ID and analyze the distribution further groupByUser in JQL.
Let me know if you need a specific example or code snippet for JQL!
kapa.ai is still learning and improving, please let me know how I did by reacting below.
ℹ️ Tip: Mention @kapa.ai in the thread for follow-ups or DM the app for 1-to-1 help.
Let me check the knowledge base for you again. One moment! 📚
Yes, you can create a table in Mixpanel Insights that shows the distribution of "id" frequencies grouped into buckets, along with the percentage of IDs in each bucket. Here’s how you can approach this:
1. Breakdown by "id": Start by breaking down your "pageview X" event by the "id" property. This will give you the count of events per "id".
2. Custom Bucketing: Since you have a high-cardinality property ("id" with 100k+ values), Mixpanel will automatically bucket the frequencies into ranges. You can customize these buckets (e.g., 0-99, 100-199, etc.) using the "Custom Buckets" feature in the Insights report. To do this, open the custom buckets modal from the overflow menu and set your desired bucket ranges. This will group the "id" values by how often they occur, according to your specified frequency ranges Numeric property Use Cases Insights: Advanced.
3. Table Visualization: Change the visualization type to "Table" in Insights. This will display each frequency bucket as a row.
4. Percentage Calculation: Mixpanel’s table will show the count of "id"s in each frequency bucket. You can also add a "percentage" column, which shows the percentage of total "id"s that fall into each frequency bucket. This is a built-in feature when using custom buckets in Mixpanel tables Insights Table Chart Updates.
Summary Table Example:
This approach will let you see, for example, what percentage of "id"s occurred 0-99 times, 100-199 times, etc., directly in a table format in Mixpanel Insights.
For more details, see the documentation on custom buckets and table chart updates.
kapa.ai is still learning and improving, please let me know how I did by reacting below.
ℹ️ Tip: Mention @kapa.ai in the thread for follow-ups or DM the app for 1-to-1 help.
Let me check the knowledge base for you again. One moment! 📚
To access the custom buckets modal from the overflow menu in Mixpanel:
1. Add a Breakdown: In your report (such as Insights), add a breakdown by the property you want to bucket (e.g., "id"). 2. Open the Overflow Menu: Next to the property breakdown, look for the overflow menu (usually represented by three dots). 3. Select "Customize Buckets": In the overflow menu, select the Customize Buckets option. This will open the custom buckets modal, where you can define your desired bucket ranges and apply them to your analysis.
This process allows you to group your property values into meaningful segments directly within the UI Custom Buckets.
kapa.ai is still learning and improving, please let me know how I did by reacting below.
ℹ️ Tip: Mention @kapa.ai in the thread for follow-ups or DM the app for 1-to-1 help.
Let me check the knowledge base for you again. One moment! 📚
Thank you for clarifying. You want to bucket "id"s based on the frequency of their occurrence (i.e., how many times each "id" appears in "pageview X" events), not by the value of the "id" itself.
Based on the knowledge sources, Mixpanel’s UI allows you to break down by a property (like "id") and then use custom buckets to group by the property’s value. However, there is no direct UI feature to bucket by the frequency of occurrence of each "id" (i.e., to group "id"s by how many times they occurred, such as 0–99, 100–199, etc.) in the standard Insights report.
Custom Buckets in Mixpanel are designed to group property values themselves (e.g., age ranges, country groups), not to bucket by the count of how often each property value appears Custom Buckets. The documentation does not mention a built-in way to create buckets based on the frequency of property values within the UI.
What you can do in the UI: - You can break down "pageview X" by "id" to see the count for each "id". - You can use custom buckets to group "id" values (e.g., by ranges of "id" values), but not by their frequency of occurrence.
If you need to bucket by frequency: The knowledge sources do not explicitly mention a way to do this in the UI. This type of analysis (bucketing by frequency of occurrence) would typically require exporting the data or using a more advanced query language like JQL, where you could: 1. Count the number of times each "id" appears. 2. Bucket those counts into ranges (e.g., 0–99, 100–199, etc.).
However, the provided sources do not include a JQL example or confirm that this exact workflow is available out-of-the-box.
Summary: I'm sorry, but the knowledge sources do not describe a way to bucket "id"s by their frequency of occurrence within the Mixpanel UI using custom buckets. This type of analysis may require exporting the data or using JQL, but the sources do not provide explicit instructions for this scenario.
kapa.ai is still learning and improving, please let me know how I did by reacting below.
ℹ️ Tip: Mention @kapa.ai in the thread for follow-ups or DM the app for 1-to-1 help.
Let me check the knowledge base for you again. One moment! 📚
Based on the knowledge sources, Mixpanel’s Custom Buckets feature allows you to group property values (such as "id") into meaningful segments for analysis. However, the documentation describes grouping by selecting specific values or ranges (for numeric properties), not by applying a substring or pattern (such as the first two characters of a string property) directly in the UI.
You can manually create a custom bucket for each prefix (e.g., "MD", "VA") by selecting all "id" values that start with those prefixes, but this would require you to individually select each value, which is not practical for large numbers of unique prefixes. The documentation does not mention a way to automatically group string property values by a substring or pattern in the UI Custom Buckets.
If you need to group "id"s by their prefix (first two characters), you would typically need to create a new property (e.g., "id_prefix") during data ingestion or via a custom property, and then use that property for your breakdowns and custom buckets. The documentation suggests creating a custom property for reusable or more complex segmentations.
Summary: I'm sorry, but the knowledge sources do not describe a way to automatically bucket by the first two characters of a property value in the Mixpanel UI. For this use case, you would need to create a new property representing the prefix and use that for your analysis.
kapa.ai is still learning and improving, please let me know how I did by reacting below.
ℹ️ Tip: Mention @kapa.ai in the thread for follow-ups or DM the app for 1-to-1 help.
