Table

The table widget renders query results as a sortable data grid. Use it when you need to display multiple columns or when a chart does not adequately represent the data.

Use it for:

• Detailed breakdowns with several dimensions
• Top N lists with multiple attributes
• Raw event inspection
• Data exploration before building a chart

---

Query requirements

A table widget accepts any SELECT that returns one or more columns. There are no restrictions on the number of columns or rows (within reasonable limits).

For simple queries, use simple mode:

Select: platform AS "Platform", count() AS "Total Records"
Filters: WHERE name = 'app_started' GROUP BY platform ORDER BY "Total Records" DESC

For multi-level aggregations that need a subquery, use full query mode:

SELECT identity, country, app_version,
       count() AS sessions,
       round(avg(duration), 1) AS avg_duration_min
FROM (
    SELECT identity, country, app_version, session_id,
           dateDiff('minute', min(timestamp), max(timestamp)) AS duration
    FROM {table}
    WHERE name = 'app_started'
    GROUP BY identity, country, app_version, session_id
)
GROUP BY identity, country, app_version
ORDER BY sessions DESC
LIMIT 100

---

Inspecting JSON fields

The table widget is particularly useful for exploring the contents of the value and custom JSON columns before you know their internal structure.

Select: value, custom
Filters: WHERE name = 'level_completed' LIMIT 10

Once you can see the field names, access them with dot notation in subsequent queries:

Select:

toString(value.level) AS level,
toString(value.difficulty) AS difficulty,
count() AS completions

Filters: WHERE name = 'level_completed' GROUP BY level, difficulty ORDER BY completions DESC

---

Common examples

Platforms breakdown (the default widget created with every project)

Select: platform AS "Platform", COUNT(platform) AS "Total Records"
Filters: WHERE name = 'app_started' GROUP BY platform ORDER BY "Total Records" DESC

---

Top 20 users by session count

Select: identity, country, uniqExact(session_id) AS sessions
Filters: WHERE name = 'app_started' GROUP BY identity, country ORDER BY sessions DESC LIMIT 20

---

Recent events

Select: timestamp, name, identity, platform, app_version
Filters: ORDER BY timestamp DESC LIMIT 50