SPSS AnswerNet: Result Solution ID: 100000681 Product: SPSS Base Title: Mixed File With No Record Identifier Description: Q. I have a file which is basically a mixed file type, but I don't have any field which identifies what type of record is on a given line of data. I was, however, keen enough to put a code which indicates when a new case is starting. Is there any way I can get SPSS to read the data? A. Yes, there is. Study the following program and modify it to fit your problem. Here are some points to consider: Most importantly: The file does not have any record type information coded in the data. Such information has to be inferred by the following information: (A) There is a means of detecting that a new case is starting (/// in first 3 columns). (B) There is a constant number of initial records (One in this case, it is easy to modify the program to permit more) (C) There are one or more lower level records each with the same structure. If there is more than one type of lower level record the person is out of luck. ********************************************************************. * This input program will read the following form of complex file * * A new case begins with a flag (in this case ///). * * The first record is a header record describing the observation * * Unit. Following that is an unknown number of records of a given * * type (in this case consisting of 3 variables). The goal is to * * construct a single record containing all variables for each case * ********************************************************************. INPUT PROGRAM. NUMERIC ID. VECTOR DATA(30). DATA LIST /#S 1-3 (A). /************************/ DO IF (#S = '///'). /* New subject begins */ + COMPUTE #COUNT=0. /* Initialize Counter */ + DATA LIST /ID 1-3 H1 TO H10 5-24. /* Read header record */ + ELSE. /* Either H or M record */ + COMPUTE #COUNT = #COUNT+1. /* Increment Counter */ + LEAVE ID H1 TO H10. /* Carry Header Info */ + REREAD. /************************/ + DATA LIST /#M1 TO #M3 1-12. /* Read an M Record */ + VECTOR #M= #M1 TO #M3. /* Utilize vectors and */ + Compute #Offset = (#COUNT-1)*3. /* Loop to Distribute */ + LOOP #X=1 to 3. /* The record into the */ + COMPUTE DATA(#Offset +#X) = #M(#X). /* Correct Locations */ + END LOOP. /************************/ END IF. END INPUT PROGRAM. BEGIN DATA. /// 001 11121314151617181910 111 112 113 121 122 123 131 132 133 141 142 143 151 152 153 161 162 163 /// 002 21222324252627282920 211 212 213 221 222 223 231 232 233 241 242 243 /// 003 31323334353637383930 311 312 313 321 322 323 331 332 333 341 342 343 351 352 353 361 362 363 371 372 373 381 382 383 391 392 393 301 302 303 END DATA. SELECT IF (NOT (SYSMIS(ID))). AGGREGATE OUTFILE= * / BREAK = ID / HEADER1 TO HEADER10 = FIRST(H1 TO H10) / DATA1 TO DATA30 = MAX ( DATA1 TO DATA30). FORMATS ALL (F3.0). LIST. THE VARIABLES ARE LISTED IN THE FOLLOWING ORDER: LINE 1: ID HEADER1 HEADER2 HEADER3 HEADER4 HEADER5 HEADER6 HEADER7 HEADER8 HEADER9 HEADER10 DATA1 DATA2 DATA3 DATA4 DATA5 DATA6 LINE 2: DATA7 DATA8 DATA9 DATA10 DATA11 DATA12 DATA13 DATA14 DATA15 DATA16 DATA17 DATA18 DATA19 DATA20 DATA21 DATA22 DATA23 LINE 3: DATA24 DATA25 DATA26 DATA27 DATA28 DATA29 DATA30 ID: 1 11 12 13 14 15 16 17 18 19 10 111 112 113 121 122 123 DATA7: 131 132 133 141 142 143 151 152 153 161 162 163 . . . . . DATA24: . . . . . . . ID: 2 21 22 23 24 25 26 27 28 29 20 211 212 213 221 222 223 DATA7: 231 232 233 241 242 243 . . . . . . . . . . . DATA24: . . . . . . . ID: 3 31 32 33 34 35 36 37 38 39 30 311 312 313 321 322 323 DATA7: 331 332 333 341 342 343 351 352 353 361 362 363 371 372 373 381 382 DATA24: 383 391 392 393 301 302 303 Number of cases read: 3 Number of cases listed: 3