Numeric Types

About

Numeric data types are used to store numbers in a database. These numbers can be integers (whole numbers) or decimals (numbers with a fractional part). SQL provides a set of standard numeric types that most relational databases support, though exact names and behavior may vary slightly between vendors.

1. INTEGER (or INT)

• Stores whole numbers without a fractional component.

• Example: 1, -20, 999

• Storage size and range can vary depending on the system.

• Often used for IDs, counters, or quantities.

Synonyms: INT, INTEGER

2. SMALLINT

• Stores smaller-range integers than INT.

• Takes less storage.

• Suitable when the range of values is limited (e.g., 0–100).

3. BIGINT

• Stores very large integers.

• Useful for applications that require large numeric values (e.g., timestamps, large counters).

4. DECIMAL(p, s) or NUMERIC(p, s)

• Stores exact numbers with a fixed number of digits.

• p = total number of digits (precision).

• s = number of digits to the right of the decimal point (scale).

• Commonly used for monetary or financial data where precision is critical.

Examples:

• DECIMAL(5,2) can store values like 123.45, 99.99, etc.

• NUMERIC(10,0) is an integer with up to 10 digits.

Note: DECIMAL and NUMERIC are functionally equivalent in standard SQL.

5. FLOAT(p)

• Stores approximate numeric values with floating-point precision.

• p is precision, typically the number of bits used to represent the value.

• Useful for scientific calculations where precision is not exact.

6. REAL

• A floating-point number with lower precision than FLOAT.

• Approximate numeric data type.

• Not suitable when exact values are required.

7. DOUBLE PRECISION

• A floating-point number with higher precision than REAL.

• Also an approximate type.

• Used when more accuracy is needed in floating-point calculations.

Vendor Support Overview

Data Type	Standard SQL	Oracle	MySQL	PostgreSQL	SQL Server
INTEGER / INT	Yes	Yes	Yes	Yes	Yes
SMALLINT	Yes	Yes	Yes	Yes	Yes
BIGINT	Yes	Yes (from 8i)	Yes	Yes	Yes
DECIMAL(p,s)	Yes	Yes	Yes	Yes	Yes
NUMERIC(p,s)	Yes	Yes	Yes	Yes	Yes
FLOAT(p)	Yes	Yes (alias)	Yes	Yes	Yes
REAL	Yes	No (not standard)	Yes	Yes	Yes
DOUBLE PRECISION	Yes	No (use BINARY_DOUBLE)	Yes	Yes	Yes (FLOAT(53))