SPSS Data Formats

In addition to the inherent type of data, SPSS stores both a print format and a write format. Both of these are forms of output formats.

Within the context of data reading and data exporting, additional input formats and output formats are allowed.

Contents

SPSS Data Formats

Print Format

The print format is used for SPSS procedures and the PRINT command. Note that w is the field width and d is the number of visible decimal places.

Format	Specification	Example
String	`Aw`	`A` as `A1` prints `A`
Hexadecimal codepoints	`AHEXw`	`A` as `AHEX2` prints `41`
Numeric	`Fw.d`	`1.234` as `F5.2` prints `1.23`
Restricted numeric	`Nw`	`123` as `N5` prints `00123`
Scientific notation	`Ew.d`	`1234` as `E10.2` prints `1234E+03`
Numeric with dollar sign	`DOLLARw.d`	`1234.56` as `DOLLAR9.2` prints `$1,234.56`
Numeric with commas	`COMMAw.d`	`1234.56` as `COMMA9.2` prints `1,234.56`
Numeric with dots	`DOTw.d`	`1234.56` as `DOT9.2` prints `1.234,56`
Numeric with percent sign	`PCTw.d`	`1234.56` as `PCT9.2` prints `1234.56%`

Date and time formats will print very differently depending on the field width. There are also limited valid widths for these fields.

Date or Time Format	Specification	Prototype
Date and time	`DATETIME20`	`dd-MMM-yyyy hh:mm:ss`
Date and time	`DATETIME17`	`dd-MMM-yyyy hh:mm`
Date	`DATE11`	`dd-MMM-yyyy`
Date	`DATE9`	`dd-MMM-yy`
Time	`TIME8`	`hh:mm:ss`
Time	`TIME5`	`hh:mm`
American date	`ADATE10`	`mm/dd/yyyy`
European date	`EDATE10`	`dd.mm.yyyy`
Sorted date	`SDATE10`	`yyyy/mm/dd`

Note that MMM prints as JAN, not 001.

Summary Formats

The following formats are only allowed on certain summary tables.

Format	Specification	Example
Parenthesize negative values	`NEGPARENw.d`	`-1234.567` as `NEGPAREN9.2` prints `(1234.57)`
Prefix with `N=`	`NEQUALw.d`	`1234.567` as `NEQUAL9.2` prints `N=1234.57`
Parenthesize values	`PARENw.d`	`1234.567` as `PAREN8.2` prints `(1234.57)`
Parenthesize percentage	`PCTPARENw.d`	`1234.567` as `PCTPAREN10.2` prints `(1234.57%)`

Write Format

The write format exactly mirrors the print format, but it is instead used by the WRITE command.

Input and Output Formats

SPSS can read and export data according to columnar or Fortran specifications. Generally speaking, these formats look like (FORMAT,d) and (FORMATw.d) respectively. Note that w is the field width and d is the implied number of decimal places.

Note however that, for free-field data input, the width and implied decimal specifications are ignored.

Columnar and Fortran Formats

Many common input formats can be expressed in columnar or Fortran styles. This is a crosswalk of the equivalent styles.

Format	Columnar Specification	Fortran Specification
Numeric	`d` or `F,d`	`Fw.d`
Restricted numeric	`N,d`	`Nw.d`
Scientific notation	`E,d`	`Ew.d`
Numeric with commas	`COMMA,d`	`COMMAw.d`
Numeric with dots	`DOT,d`	`DOTw.d`
Numeric with dollar sign	`DOLLAR,d`	`DOLLARw.d`
Numeric with percent sign	`PCT,d`	`PCTw.d`
Zoned decimal	`Z,d`	`Zw.d`
String	`A`	`Aw`

For columnar specifications, the numeric format is the default and needs no specification.

When the restricted numeric format is used for fixed-width input, if the input field has any other width, the value is read as system missing. For either fixed-width or free-field input, if the input field has a sign (-) or a decimal point (.), the value is read as system missing.

Fortran Formats

The Fortran style offers additional instructions and formats.

Instruction	Fortran Specification
Move to `n`th column	`Tn`
Repeat the following specification `n` times	`nSPEC`

Date or Time Prototype	Fortran Specification
`dd-mmm-yyyy`	`DATEw`
`mm/dd/yyyy`	`ADATEw`
`dd/mm/yy`	`EDATEw`
`yyddd`	`JDATEw`
`yy/mm/dd`	`SDATEw`
`qQyyyy`	`QYRw`
`mm/yyyy`	`MOYRw`
`wkWKyyyy`	`WKYRw`
`dd-mmm-yyyy hh:mm:ss.ss`	`DATETIMEw.d`
`hh:mm:ss.ss`	`TIMEw.d`
`ddd hh:mm:ss.ss`	`DTIMEw.d`
	`WKDAYw`
	`MONTHw`

Note that SPSS is capable of reading date and time data in columnar specifications. Simply use the standard date and time formats, as seen above under print formats.

Binary Formats

The following formats are used for reading data from COBOL, Fortran, or other low-level programs.

Format	Specification	Example
Fixed point integer binary	`IBw.d`	`11111111` as `IB8.0` reads `-1`
Positive fixed point integer binary	`PIBw.d`	`11111111` as `IB8.0` reads `255`
Positive fixed point integer hexadecimal	`PIBHEXw`	`FFFF` as `PIBHEX4` reads `65535`
Zoned decimal	`Zw.d`	`1234A` as `Z5.2` reads `123.41`
Packed decimal	`Pw.d`
Unsigned packed decimal	`PKw.d`
Real binary	`RBw`
Real hexadecimal	`RBHEXw`

The fixed point integer binary format is modeled after the data used by COBOL in COMPUTATIONAL data items, or by FORTRAN in INTEGER*2 or INTEGER*4, or by Assembler in fullword and halfword items.

The zoned decimal format is modeled after the data used by COBOL in DISPLAY data items, or by PL/I in PICTURE data items, or by Assembler in zoned decimal data items.

The packed decimal format is modeled after the data used by COBOL in COMPUTATIONAL–3 data items, or by Assembler in packed decimal data items.

The real binary format is modeled after the data used by COBOL in COMPUTATIONAL–1 or COMPUTATIONAL–2 data items, or by PL/I in FLOATING DECIMAL data items, or by FORTRAN in REAL or REAL*8 data items, or by Assembler in floating-point data items.

After Reading Data

Once data is read into SPSS, the stored output formats are determined from the input format.

Implied decimal spaces are converted into actual decimal places, and 1 is added to the format width. For example, 1234 with an input format of F5.2 is read as 12.34 with an output format of F6.2.

Decorations and indicators are added to the format width. For example, data read with an input format DOLLAR7.2 receives an output format of DOLLAR10.2 to fit the dollar sign ($), comma (,), and decimmal (.).

Binary formats all become the standard numeric format.

CategoryRicottone