Data Types

Singdata Lakehouse supports numeric, string, time, boolean, binary, and complex types (ARRAY, MAP, STRUCT, JSON, VECTOR, BITMAP). It is compatible with common type aliases from MySQL, PostgreSQL, Hive, and other systems, so migrated DDL scripts can run without manually replacing type names.

Type Quick Reference

Numeric Types

Type	Storage	Use Case
`TINYINT`	1 byte	Status codes, enum values
`SMALLINT`	2 bytes	Year values, small-range counters
`INT`	4 bytes	General integer IDs, counters
`BIGINT`	8 bytes	Large-scale IDs, millisecond timestamps
`FLOAT`	4 bytes	Low-precision floating point, scientific computing
`DOUBLE`	8 bytes	High-precision floating point, statistical analysis
`DECIMAL(p,s)`	Variable	Monetary amounts, exact calculations (recommended)

String Types

Type	Length	Use Case
`CHAR(n)`	Fixed, max 255	Fixed-length codes (country codes, status codes)
`VARCHAR(n)`	Variable, max 1,048,576	Fields with a length cap (usernames, email addresses)
`STRING`	Unlimited	Long text, JSON strings (recommended)

⚠️ Note: All three string types silently truncate when the declared length is exceeded — no error is raised.

Time Types

Type	Timezone	Use Case
`DATE`	None	Date dimensions, birthdays
`TIMESTAMP`	With timezone (stored as UTC)	Cross-timezone event times, log timestamps
`TIMESTAMP_NTZ`	No timezone (stored as-is)	Local business time, no timezone conversion needed
`INTERVAL`	—	Time differences, date arithmetic

Other Basic Types

Type	Description
`BOOLEAN`	`TRUE` / `FALSE` / `NULL`
`BINARY`	Binary byte sequence

Complex Types

Type	Description	Use Case
`ARRAY<T>`	Ordered collection of same-type elements	Tag lists, multi-value attributes
`MAP<K,V>`	Key-value pair collection	Dynamic attributes, configuration items
`STRUCT<...>`	Nested structure with fixed fields	Addresses, coordinates, and other structured nesting
`JSON`	Semi-structured JSON document	Flexible schema, event attributes
`VECTOR(n)`	Fixed-dimension floating-point vector	Embedding storage, vector search, RAG
`BITMAP`	Compressed integer set (Roaring Bitmap)	UV statistics, user tag segmentation

In This Chapter

Page	Description
Data Types Reference	Full description of each type, type alias mapping, migration compatibility
Data Type Conversion	Explicit conversion (CAST / `::`) and implicit conversion rules
Numeric Types	TINYINT / SMALLINT / INT / BIGINT / FLOAT / DOUBLE / DECIMAL in detail
String Types	CHAR / VARCHAR / STRING comparison and selection guide
Date/Time Types	DATE / TIMESTAMP / TIMESTAMP_NTZ / INTERVAL comparison and selection guide
BINARY	Binary type reference
BOOLEAN	Boolean type reference
ARRAY	Array type syntax and functions
MAP	Map type syntax and functions
STRUCT	Nested struct type syntax and functions
JSON	JSON type syntax, field access, and query examples
VECTOR	Vector type syntax, dimension configuration, and vector index usage
BITMAP	Bitmap type syntax, set operation functions reference

Document	Description
Data Type Conversion Guide	Type mapping when migrating from MySQL / PostgreSQL / Hive
TIMESTAMP_NTZ Reference	Detailed explanation and use cases for timezone-free timestamps
CREATE TABLE	Full syntax for using data types when creating tables