Conversion rules in Standard SQL

"Conversion" includes, but is not limited to, casting and coercion.

Casting is explicit conversion and uses the CAST() function.
Coercion is implicit conversion, which BigQuery performs automatically under the conditions described below.
There is a third group of conversion functions that have their own function names, such as UNIX_DATE().

The table below summarizes all possible CAST and coercion possibilities for BigQuery data types. "Coercion To" applies to all expressions of a given data type (e.g. a column) , but literals and parameters can also be coerced. See Literal Coercion and Parameter Coercion for details.

From Type	CAST to	Coercion To
INT64	BOOL INT64 NUMERIC FLOAT64 STRING	FLOAT64 NUMERIC
NUMERIC	INT64 NUMERIC FLOAT64 STRING	FLOAT64
FLOAT64	INT64 NUMERIC FLOAT64 STRING
BOOL	BOOL INT64 STRING
STRING	BOOL INT64 NUMERIC FLOAT64 STRING BYTES DATE DATETIME TIME TIMESTAMP
BYTES	BYTES STRING
DATE	DATE DATETIME STRING TIMESTAMP
DATETIME	DATE DATETIME STRING TIME TIMESTAMP
TIME	STRING TIME
TIMESTAMP	DATE DATETIME STRING TIME TIMESTAMP
ARRAY	ARRAY
STRUCT	STRUCT

Casting

Syntax:

CAST(expr AS typename)

Cast syntax is used in a query to indicate that the result type of an expression should be converted to some other type.

Example:

CAST(x=1 AS STRING)

This results in "true" if x is 1, "false" for any other non-NULL value, and NULL if x is NULL.

Casts between supported types that do not successfully map from the original value to the target domain produce runtime errors. For example, casting BYTES to STRING where the byte sequence is not valid UTF-8 results in a runtime error.

When casting an expression x of the following types, these rules apply:

From	To	Rule(s) when casting `x`
INT64	FLOAT64	Returns a close but potentially not exact FLOAT64 value.
INT64	BOOL	Returns `FALSE` if `x` is `0`, `TRUE` otherwise.
NUMERIC	Floating Point	NUMERIC will convert to the closest floating point number with a possible loss of precision.
FLOAT64	INT64	Returns the closest INT64 value. Halfway cases such as 1.5 or -0.5 round away from zero.
FLOAT64	STRING	Returns an approximate string representation.
FLOAT64	NUMERIC	If the floating point number has more than nine digits after the decimal point, it will round half away from zero. Casting a `NaN`, `+inf` or `-inf` will return an error. Casting a value outside the range of `NUMERIC` will return an overflow error.
BOOL	INT64	Returns `1` if `x` is `TRUE`, `0` otherwise.
BOOL	STRING	Returns `"true"` if `x` is `TRUE`, `"false"` otherwise.
STRING	FLOAT64	Returns `x` as a FLOAT64 value, interpreting it as having the same form as a valid FLOAT64 literal. Also supports casts from `"inf"`, `"+inf"`, `"-inf"`, and `"nan"`. Conversions are case-insensitive.
STRING	NUMERIC	The numeric literal contained in the `STRING` must not exceed the maximum precision or range of the `NUMERIC` type, or an error will occur. If the number of digits after the decimal point exceeds nine, then the resulting `NUMERIC` value will round half away from zero to have nine digits after the decimal point.
STRING	BOOL	Returns `TRUE` if `x` is `"true"` and `FALSE` if `x` is `"false"` All other values of `x` are invalid and throw an error instead of casting to BOOL. STRINGs are case-insensitive when converting to BOOL.
STRING	BYTES	STRINGs are cast to BYTES using UTF-8 encoding. For example, the STRING "©", when cast to BYTES, would become a 2-byte sequence with the hex values C2 and A9.
BYTES	STRING	Returns `x` interpreted as a UTF-8 STRING. For example, the BYTES literal `b'\xc2\xa9'`, when cast to STRING, is interpreted as UTF-8 and becomes the unicode character "©". An error occurs if `x` is not valid UTF-8.
ARRAY	ARRAY	Must be the exact same ARRAY type.
STRUCT	STRUCT	Allowed if the following conditions are met: The two STRUCTs have the same number of fields. The original STRUCT field types can be explicitly cast to the corresponding target STRUCT field types (as defined by field order, not field name).

Safe casting

When using CAST, a query can fail if BigQuery is unable to perform the cast. For example, the following query generates an error:

SELECT CAST("apple" AS INT64) AS not_a_number;

If you want to protect your queries from these types of errors, you can use SAFE_CAST. SAFE_CAST is identical to CAST, except it returns NULL instead of raising an error.

SELECT SAFE_CAST("apple" AS INT64) AS not_a_number;

+--------------+
| not_a_number |
+--------------+
| NULL         |
+--------------+

If you are casting from bytes to strings, you can also use the function, SAFE_CONVERT_BYTES_TO_STRING. Any invalid UTF-8 characters are replaced with the unicode replacement character, U+FFFD. See SAFE_CONVERT_BYTES_TO_STRING for more information.

Casting hex strings to integers

If you are working with hex strings (0x123), you can cast those strings as integers:

SELECT '0x123' as hex_value, CAST('0x123' as INT64) as hex_to_int;

+-----------+------------+
| hex_value | hex_to_int |
+-----------+------------+
| 0x123     | 291        |
+-----------+------------+

SELECT '-0x123' as hex_value, CAST('-0x123' as INT64) as hex_to_int;

+-----------+------------+
| hex_value | hex_to_int |
+-----------+------------+
| -0x123    | -291       |
+-----------+------------+

Casting date types

BigQuery supports casting date types to/from strings as follows:

CAST(date_expression AS STRING)
CAST(string_expression AS DATE)

Casting from a date type to a string is independent of time zone and is of the form YYYY-MM-DD. When casting from string to date, the string must conform to the supported date literal format, and is independent of time zone. If the string expression is invalid or represents a date that is outside of the supported min/max range, then an error is produced.

Casting timestamp types

BigQuery supports casting timestamp types to/from strings as follows:

CAST(timestamp_expression AS STRING)
CAST(string_expression AS TIMESTAMP)

When casting from timestamp types to string, the timestamp is interpreted using the default time zone, UTC. The number of subsecond digits produced depends on the number of trailing zeroes in the subsecond part: the CAST function will truncate zero, three, or six digits.

When casting from string to a timestamp, string_expression must conform to the supported timestamp literal formats, or else a runtime error occurs. The string_expression may itself contain a time_zone—see time zones. If there is a time zone in the string_expression, that time zone is used for conversion, otherwise the default time zone, UTC, is used. If the string has fewer than six digits, then it is implicitly widened.

An error is produced if the string_expression is invalid, has more than six subsecond digits (i.e. precision greater than microseconds), or represents a time outside of the supported timestamp range.

Casting between date and timestamp types

BigQuery supports casting between date and timestamp types as follows:

CAST(date_expression AS TIMESTAMP)
CAST(timestamp_expression AS DATE)

Casting from a date to a timestamp interprets date_expression as of midnight (start of the day) in the default time zone, UTC. Casting from a timestamp to date effectively truncates the timestamp as of the default time zone.

Coercion

BigQuery coerces the result type of an expression to another type if needed to match function signatures. For example, if function func() is defined to take a single argument of type INT64 and an expression is used as an argument that has a result type of FLOAT64, then the result of the expression will be coerced to INT64 type before func() is computed.

Literal coercion

BigQuery supports the following literal coercions:

Input Data Type	Result Data Type	Notes
STRING literal	DATE TIMESTAMP

Literal coercion is needed when the actual literal type is different from the type expected by the function in question. For example, if function func() takes a DATE argument, then the expression func("2014-09-27") is valid because the STRING literal "2014-09-27" is coerced to DATE.

Literal conversion is evaluated at analysis time, and gives an error if the input literal cannot be converted successfully to the target type.

Note: String literals do not coerce to numeric types.

Parameter coercion

BigQuery supports the following parameter coercions:

Input Data Type	Result Data Type
STRING parameter

If the parameter value cannot be coerced successfully to the target type, an error is provided.

Additional conversion functions

BigQuery provides the following additional conversion functions: