MISSING Subcommand (RECORD TYPE command)

MISSING controls whether the program issues a warning when it encounters a missing record type for a case. Regardless of whether the program issues the warning, it builds the case in the active dataset with system-missing values for the variables defined on the missing record.

  • The only specification is a single keyword. NOWARN is the default for nested files. WARN is the default for grouped files. MISSING cannot be used with MIXED files.
  • MISSING on RECORD TYPE applies only to those records specified by that RECORD TYPE command. The treatment of missing records for record types without the MISSING specification on RECORD TYPE is determined by the MISSING subcommand on FILE TYPE.
  • For grouped files, the program checks whether there is a record for each case identification number. For nested files, the program verifies that each defined case includes one record of each type.

WARN. Issue a warning message when a record type is missing for a case. This is the default for grouped files.

NOWARN. Suppress the warning message when a record type is missing for a case. This is the default for nested files.

Example

FILE TYPE GROUPED FILE=HUBDATA RECORD=#RECID 80 CASE=ID 1-5.
RECORD TYPE 1.
DATA LIST   /MOHIRED YRHIRED 12-15 DEPT79 TO DEPT82 SEX 16-20.
RECORD TYPE 2  MISSING=NOWARN.
DATA LIST   /SALARY79 TO SALARY82 6-25 
  HOURLY81 HOURLY82 40-53 (2) PROMO81 72 AGE 54-55 RAISE82 66-70.
RECORD TYPE 3.
DATA LIST   /JOBCAT 6 NAME 25-48 (A).
END FILE TYPE.
  • MISSING is not specified on FILE TYPE. Therefore the default MISSING=WARN is in effect for all record types.
  • MISSING=NOWARN is specified on the second RECORD TYPE, overriding the default setting for type 2 records. WARN is still in effect for type 1 and type 3 records.