Developer Documentation Library > Data Model > Available DSCs > SPSS Statistics SAV DSC > Reading from an SPSS Statistics .sav file > Variable definitions when reading from a .sav file > IBM SPSS Statistics numeric variables
 
IBM SPSS Statistics numeric variables
By default, when generating an MDM document from the .sav file or accessing the case data with no metadata, IBM SPSS Statistics numeric variables (that are not part of an IBM SPSS Statistics multiple-response set) are mapped to MDM single-response categorical, long, or double variables according to the rules shown in the following table:
IBM SPSS Statistics variable type
MDM variable type
Notes
Numeric that has at least one value label that is not defined as a user-missing value
Single-response categorical
When generating an MDM document from the .sav file or accessing the case data with no metadata, the SPSS Statistics SAV DSC generates categories for all labeled values, including those defined as user-missing. The SPSS Statistics SAV DSC then scans the case data and generates categories for any data values that do not have a corresponding value label. However, you can stop the SPSS Statistics SAV DSC from scanning the case data and generating categories. For more information, see Determining categoriesD.
Numeric that has a decimals format of 0 (except for COMMA, DOT, and DOLLAR variable types), and that has either no value labels or only value labels that are defined as user-missing values
Long
When generating an MDM document from the .sav file or accessing the case data with no metadata, the SPSS Statistics SAV DSC identifies all IBM SPSS Statistics numeric 1.0 variables (that have no value labels or only value labels that are defined as user-missing values) as MDM long variables. To force the SPSS Statistics SAV DSC to interpret one or more of these variables as MDM boolean variables, use the NumericAsBool property (see Properties and settings used by the SPSS Statistics SAV DSC).
Handling null values
Null is returned for numeric values that are less than -2,147,483,648 or greater than 2,147,483,647.
▪To return the Double type instead, do one of the following:
▪use the ScaleAsDouble property in the generated .mdd file
▪in IBM SPSS Statistics, set the number of decimals to 1.
▪To return the Text type instead: In IBM SPSS Statistics, change the variable to “string”.
Numeric that has a decimals format of greater than 0 (except for COMMA, DOT, and DOLLAR type; these variable types are always treated as double), and that has either no value labels or only value labels that are defined as user-missing values
Double
 
Notes
Value labels are case sensitive
To change how the SPSS Statistics SAV DSC determines if a numeric variable is categorical, see Determining categorical variables.
If the IBM SPSS Statistics variable's value is the system-missing value, the value of the corresponding MDM variable is set to NULL.
See also
Variable definitions when reading from a .sav file