8.1. Numeric Types

Numeric types consist of two-, four-, and eight-byte integers, four- and eight-byte floating-point numbers, and selectable-precision decimals. Table 8.2 lists the available types.

Table 8.2. Numeric Types

NameStorage SizeDescriptionRange
smallint2 bytessmall-range integer-32768 to +32767
integer4 bytestypical choice for integer-2147483648 to +2147483647
bigint8 byteslarge-range integer-9223372036854775808 to +9223372036854775807
decimalvariableuser-specified precision, exactup to 131072 digits before the decimal point; up to 16383 digits after the decimal point
numericvariableuser-specified precision, exactup to 131072 digits before the decimal point; up to 16383 digits after the decimal point
real4 bytesvariable-precision, inexact6 decimal digits precision
double precision8 bytesvariable-precision, inexact15 decimal digits precision
smallserial2 bytessmall autoincrementing integer1 to 32767
serial4 bytesautoincrementing integer1 to 2147483647
bigserial8 byteslarge autoincrementing integer1 to 9223372036854775807

The syntax of constants for the numeric types is described in Section 4.1.2. The numeric types have a full set of corresponding arithmetic operators and functions. Refer to Chapter 9 for more information. The following sections describe the types in detail.

8.1.1. Integer Types

The types smallint, integer, and bigint store whole numbers, that is, numbers without fractional components, of various ranges. Attempts to store values outside of the allowed range will result in an error.

The type integer is the common choice, as it offers the best balance between range, storage size, and performance. The smallint type is generally only used if disk space is at a premium. The bigint type is designed to be used when the range of the integer type is insufficient.

SQL only specifies the integer types integer (or int), smallint, and bigint. The type names int2, int4, and int8 are extensions, which are also used by some other SQL database systems.

8.1.2. Arbitrary Precision Numbers

The type numeric can store numbers with a very large number of digits. It is especially recommended for storing monetary amounts and other quantities where exactness is required. Calculations with numeric values yield exact results where possible, e.g., addition, subtraction, multiplication. However, calculations on numeric values are very slow compared to the integer types, or to the floating-point types described in the next section.

We use the following terms below: The precision of a numeric is the total count of significant digits in the whole number, that is, the number of digits to both sides of the decimal point. The scale of a numeric is the count of decimal digits in the fractional part, to the right of the decimal point. So the number 23.5141 has a precision of 6 and a scale of 4. Integers can be considered to have a scale of zero.

Both the maximum precision and the maximum scale of a numeric column can be configured. To declare a column of type numeric use the syntax:

NUMERIC(precision, scale)

The precision must be positive, the scale zero or positive. Alternatively:

NUMERIC(precision)

selects a scale of 0. Specifying:

NUMERIC

without any precision or scale creates a column in which numeric values of any precision and scale can be stored, up to the implementation limit on precision. A column of this kind will not coerce input values to any particular scale, whereas numeric columns with a declared scale will coerce input values to that scale. (The SQL standard requires a default scale of 0, i.e., coercion to integer precision. We find this a bit useless. If you're concerned about portability, always specify the precision and scale explicitly.)

Note

The maximum allowed precision when explicitly specified in the type declaration is 1000; NUMERIC without a specified precision is subject to the limits described in Table 8.2.

If the scale of a value to be stored is greater than the declared scale of the column, the system will round the value to the specified number of fractional digits. Then, if the number of digits to the left of the decimal point exceeds the declared precision minus the declared scale, an error is raised.

Numeric values are physically stored without any extra leading or trailing zeroes. Thus, the declared precision and scale of a column are maximums, not fixed allocations. (In this sense the numeric type is more akin to varchar(n) than to char(n).) The actual storage requirement is two bytes for each group of four decimal digits, plus three to eight bytes overhead.

In addition to ordinary numeric values, the numeric type allows the special value NaN, meaning not-a-number. Any operation on NaN yields another NaN. When writing this value as a constant in an SQL command, you must put quotes around it, for example UPDATE table SET x = 'NaN'. On input, the string NaN is recognized in a case-insensitive manner.

Note

In most implementations of the not-a-number concept, NaN is not considered equal to any other numeric value (including NaN). In order to allow numeric values to be sorted and used in tree-based indexes, Postgres Pro treats NaN values as equal, and greater than all non-NaN values.

The types decimal and numeric are equivalent. Both types are part of the SQL standard.

When rounding values, the numeric type rounds ties away from zero, while (on most machines) the real and double precision types round ties to the nearest even number. For example:

SELECT x,
  round(x::numeric) AS num_round,
  round(x::double precision) AS dbl_round
FROM generate_series(-3.5, 3.5, 1) as x;
  x   | num_round | dbl_round
------+-----------+-----------
 -3.5 |        -4 |        -4
 -2.5 |        -3 |        -2
 -1.5 |        -2 |        -2
 -0.5 |        -1 |        -0
  0.5 |         1 |         0
  1.5 |         2 |         2
  2.5 |         3 |         2
  3.5 |         4 |         4
(8 rows)

8.1.3. Floating-Point Types

The data types real and double precision are inexact, variable-precision numeric types. On all currently supported platforms, these types are implementations of IEEE Standard 754 for Binary Floating-Point Arithmetic (single and double precision, respectively), to the extent that the underlying processor, operating system, and compiler support it.

Inexact means that some values cannot be converted exactly to the internal format and are stored as approximations, so that storing and retrieving a value might show slight discrepancies. Managing these errors and how they propagate through calculations is the subject of an entire branch of mathematics and computer science and will not be discussed here, except for the following points:

  • If you require exact storage and calculations (such as for monetary amounts), use the numeric type instead.

  • If you want to do complicated calculations with these types for anything important, especially if you rely on certain behavior in boundary cases (infinity, underflow), you should evaluate the implementation carefully.

  • Comparing two floating-point values for equality might not always work as expected.

On all currently supported platforms, the real type has a range of around 1E-37 to 1E+37 with a precision of at least 6 decimal digits. The double precision type has a range of around 1E-307 to 1E+308 with a precision of at least 15 digits. Values that are too large or too small will cause an error. Rounding might take place if the precision of an input number is too high. Numbers too close to zero that are not representable as distinct from zero will cause an underflow error.

By default, floating point values are output in text form in their shortest precise decimal representation; the decimal value produced is closer to the true stored binary value than to any other value representable in the same binary precision. (However, the output value is currently never exactly midway between two representable values, in order to avoid a widespread bug where input routines do not properly respect the round-to-nearest-even rule.) This value will use at most 17 significant decimal digits for float8 values, and at most 9 digits for float4 values.

Note

This shortest-precise output format is much faster to generate than the historical rounded format.

For compatibility with output generated by older versions of Postgres Pro, and to allow the output precision to be reduced, the extra_float_digits parameter can be used to select rounded decimal output instead. Setting a value of 0 restores the previous default of rounding the value to 6 (for float4) or 15 (for float8) significant decimal digits. Setting a negative value reduces the number of digits further; for example -2 would round output to 4 or 13 digits respectively.

Any value of extra_float_digits greater than 0 selects the shortest-precise format.

Note

Applications that wanted precise values have historically had to set extra_float_digits to 3 to obtain them. For maximum compatibility between versions, they should continue to do so.

In addition to ordinary numeric values, the floating-point types have several special values:


Infinity
-Infinity
NaN

These represent the IEEE 754 special values infinity, negative infinity, and not-a-number, respectively. When writing these values as constants in an SQL command, you must put quotes around them, for example UPDATE table SET x = '-Infinity'. On input, these strings are recognized in a case-insensitive manner.

Note

IEEE754 specifies that NaN should not compare equal to any other floating-point value (including NaN). In order to allow floating-point values to be sorted and used in tree-based indexes, Postgres Pro treats NaN values as equal, and greater than all non-NaN values.

Postgres Pro also supports the SQL-standard notations float and float(p) for specifying inexact numeric types. Here, p specifies the minimum acceptable precision in binary digits. Postgres Pro accepts float(1) to float(24) as selecting the real type, while float(25) to float(53) select double precision. Values of p outside the allowed range draw an error. float with no precision specified is taken to mean double precision.

8.1.4. Serial Types

Note

This section describes a Postgres Pro-specific way to create an autoincrementing column. Another way is to use the SQL-standard identity column feature, described at CREATE TABLE.

The data types smallserial, serial and bigserial are not true types, but merely a notational convenience for creating unique identifier columns (similar to the AUTO_INCREMENT property supported by some other databases). In the current implementation, specifying:

CREATE TABLE tablename (
    colname SERIAL
);

is equivalent to specifying:

CREATE SEQUENCE tablename_colname_seq AS integer;
CREATE TABLE tablename (
    colname integer NOT NULL DEFAULT nextval('tablename_colname_seq')
);
ALTER SEQUENCE tablename_colname_seq OWNED BY tablename.colname;

Thus, we have created an integer column and arranged for its default values to be assigned from a sequence generator. A NOT NULL constraint is applied to ensure that a null value cannot be inserted. (In most cases you would also want to attach a UNIQUE or PRIMARY KEY constraint to prevent duplicate values from being inserted by accident, but this is not automatic.) Lastly, the sequence is marked as owned by the column, so that it will be dropped if the column or table is dropped.

Note

Because smallserial, serial and bigserial are implemented using sequences, there may be "holes" or gaps in the sequence of values which appears in the column, even if no rows are ever deleted. A value allocated from the sequence is still "used up" even if a row containing that value is never successfully inserted into the table column. This may happen, for example, if the inserting transaction rolls back. See nextval() in Section 9.16 for details.

To insert the next value of the sequence into the serial column, specify that the serial column should be assigned its default value. This can be done either by excluding the column from the list of columns in the INSERT statement, or through the use of the DEFAULT key word.

The type names serial and serial4 are equivalent: both create integer columns. The type names bigserial and serial8 work the same way, except that they create a bigint column. bigserial should be used if you anticipate the use of more than 231 identifiers over the lifetime of the table. The type names smallserial and serial2 also work the same way, except that they create a smallint column.

The sequence created for a serial column is automatically dropped when the owning column is dropped. You can drop the sequence without dropping the column, but this will force removal of the column default expression.