Sample usage
CAST(Number_field AS TEXT)
CAST(Text_field AS NUMBER)
CAST(Date_field as DATETIME)
Syntax
Parameters
field_expression - a field or expression.
type - a data type literal.
CAST data type literals
You can cast from/to the following types:
| From type | To type |
|---|---|
| Text |
|
| Number |
|
| Date |
|
| Date & Time |
|
Examples
Goal: count website sessions longer than N seconds.
In a Google Analytics data source, Session Duration is a text dimension. The first step to counting sessions longer than N is to convert this to a number using CAST. The second step is to use the CASE statement to identify sessions longer than a certain value—we'll say 60 seconds for this example. We'll combine CASE with SUM to do the counting.
Step 1: Create a new calculated field Session Duration (number) with the following formula:
CAST(Session Duration AS NUMBER)
Step 2: Create a new calculated field Long Sessions:
SUM(CASE WHEN Session Duration > 60 THEN 1 ELSE 0 END)
Notes
You cannot CAST aggregated fields.
- Casts from
[DATE|DATETIME]toTEXTreturn the canonical string formats below. To return a different format, use theFORMAT_DATETIMEfunction. - Casts from
TEXTto[DATE|DATETIME]accept only the canonical string format. To parse a different format, use thePARSE_*functions.
| Type | Canonical STRING format |
|---|---|
| DATE | YYYY-[M]M-[D]D |
| DATETIME | YYYY-[M]M-[D]D [[H]H:[M]M:[S]S[.DDDDDD]] |
Coercion to text
When an expression requires a text string, Looker Studio coerces (implicitly converts) non-string values to a string data type.
| Type | Canonical STRING format |
|---|---|
| DATE | YYYY-MM-DD |
| DATETIME | YYYY-MM-DD HH:MM:SS[.DDD[DDD]] |
For example, suppose we have a created_datetime field of type DATETIME and the following formula:
CONCAT("Created on ", created_datetime)
This formula returns the following:
Created on 2019-11-22 16:30:00.739182
This formula doesn’t require an explicit CAST(created_datetime AS TEXT).