NO_COMMON_TYPE ClickHouse® error¶

The NO_COMMON_TYPE error in ClickHouse® (and Tinybird) happens when you try to perform operations between values of incompatible types, and ClickHouse® can't determine a common type for the operation. This commonly occurs in comparisons, arithmetic operations, or when combining different data types.

What causes this error¶

You typically see it when:

• Comparing values of different types (for example, String vs Int)
• Performing arithmetic operations on mixed types
• Using UNION with incompatible column types
• Combining columns with different data types in expressions
• Using functions that expect specific types but receive others

Example errors¶

SELECT if(user_id = 123, 'user_123', user_id) FROM events

SELECT value + '10' FROM events

SELECT user_id FROM events
UNION ALL
SELECT email FROM users

SELECT CASE
    WHEN user_id = 123 THEN 'user_123'
    WHEN user_id = 456 THEN user_id
END FROM events

How to fix it¶

Use explicit type casting¶

Convert values to compatible types using CAST or type conversion functions:

SELECT * FROM events WHERE user_id = toInt32('123')

SELECT value + toInt32('10') FROM events

Use proper type conversion functions¶

ClickHouse® provides specific conversion functions for different types:

SELECT * FROM events WHERE user_id = toInt32(user_id_string)

SELECT toString(user_id) as user_id_str FROM events

Fix UNION queries¶

Ensure all columns in UNION have compatible types:

SELECT toString(user_id) as id FROM events
UNION ALL
SELECT email as id FROM users

Common patterns and solutions¶

String vs Numeric comparisons¶

When comparing strings with numbers:

SELECT if(user_id = 123, 'user_123', user_id) FROM events

WHERE user_id = toInt32('123')

WHERE toString(user_id) = '123'

Date and DateTime operations¶

When working with dates and timestamps:

WHERE date_column = now()

WHERE toDate(date_column) = toDate(now())

Array operations¶

When working with arrays of different types:

SELECT [1, 2, 3] + ['a', 'b', 'c']

SELECT [toString(1), toString(2), toString(3)] + ['a', 'b', 'c']

Advanced solutions¶

Using CASE statements with mixed types¶

Convert all branches to a common type:

SELECT CASE
    WHEN user_id = 123 THEN 'user_123'
    WHEN user_id = 456 THEN 'user_456'
    ELSE toString(user_id)
END as user_label FROM events

Working with JSON data¶

When extracting values from JSON:

SELECT
    toInt32(JSONExtractRaw(data, 'user_id')) as user_id,
    toString(JSONExtractRaw(data, 'event_type')) as event_type
FROM events

Handling NULL values¶

Be careful with NULL values in type conversions:

SELECT toInt32OrZero(user_id_string) as user_id FROM events

Tinybird-specific notes¶

In Tinybird, this error often occurs when:

• Data Sources have inconsistent data types in the same column
• Pipes try to combine data from different sources with incompatible schemas
• Schema inference assigns wrong types to columns
• JSON data contains mixed types

To debug in Tinybird:

1. Check your Data Source schema for type consistency
2. Use Schema Hints to enforce proper types
3. Add explicit CAST operations in your Pipes
4. Validate incoming data types in Data Sources

Type conversion functions¶

Common ClickHouse® type conversion functions:

• toInt8(), toInt16(), toInt32(), toInt64() - Convert to integers
• toUInt8(), toUInt16(), toUInt32(), toUInt64() - Convert to unsigned integers
• toFloat32(), toFloat64() - Convert to floating point
• toString() - Convert to string
• toDate(), toDateTime() - Convert to date/time types
• toUUID() - Convert to UUID

See also¶

• ClickHouse® Data Types
• Type Conversion Functions
• Common function errors