Fixed length data file with multiple records per line

Data files transferred to PC's from Mainframe computers are frequently in a fixed-length format with no "Carriage Return - Line Feed" indicators at the end of the records. For these files, each record contains a fixed number of characters, its "record length".

 

Here is such a file. I have broken it into lines of twice the "record length" for the sake of convenient display, but the file itself is just one long string of numbers.

 

This is data file, TAPEDAT.txt:

 

Fred    2625232032131000110110George  2925202423320100110111

Mabel   2622212643411001000003Alice   2923212421321010010011

Grace   2723262424311111011012Arthur  2523222222140001010111

Mary    2825212412431100000003Belinda 2627222431320111011111

 

The test is of 10 items, divided into 4 groups. Group 1 is items 1,2,3. Group 2 is items 4,5,6. Group 3 is items 7,8,9. Group 4 is just item 10. The judging plan rotates the judges across subjects, item groups and sessions in a random manner, but without duplication of rating. Thus the items in group one for subject Fred were rated by judge 26 in session 3. Item 1 was rated 1, Items 2 and 3 were rated 0. The first 9 items are dichotomies (0=failed, 1= passed). Item 10 is on a four-point Likert scale scored 0, 1, 2 ,3. Schematically the data record is:

 

This is entered in each 30 character record as follows:

 

Columns

 1-8 Subject

9-10 Judge for item group 1

11-12 Judge for item group 2

13-14 Judge for item group 3

15-16 Judge for item group 4

  17 Session number for item group 1

  18 Session number for item group 2

  19 Session number for item group 3

  20 Session number for item group 4

21-30 Ratings for Items 1-10

 

The Facform keyword file, TAPEkey.txt, to reformat this fixed length file is:

 

; File: TAPEkey.txt

; This converts a file with multiple records per data line

$Input = tapeDAT.txt

$Length = 30         ; the fixed length record length

$Output = tapeFAC.txt

 

$Facets = 5 ;Subject, Item group, Judge, Session, Item

$Flabel = 1, Subject

$Flabel = 2, "Item Group"  ; This facet is included "just in case"

$Flabel = 3, Judge

$Flabel = 4, Session

$Flabel = 5, Item

 

$Spec = "Title = Demonstration of Fixed Length data file"

$Spec = "Output = tapeOUT.txt ; the output of the FACETS analysis"

$Spec = "Models =       ; we model only subject-judge-item here"

$Spec = "?,,?,,10,R     ;Item 10 has a rating (or partial credit) scale"

$Spec = "?,,?,,?,D      ;Items 1-9 dichotomies"

$Spec = "*"

 

$Label = 1,,$S1W8        ;Subject name

Ig = 0  ;used to keep track of item group

Item = 0 ; reset item number

$Do = 4

Ig = Ig +1

$Label = 2,Ig        ;Item group

$Label = 3,$S(IG*2+7)W2 ;Judge

$Label = 4,$S(IG+16)W1  ;Session

$Do = 3 ;Items in groups of 3

Item = Item +1

$If = (Item=<10)     ;No item greater than 10

$Label = 5,Item,Item   ;Item

$Rating = $S(Item+20)W1 ;Rating

 $Endif

$Again

$Again

 

The resulting Facets file is TAPEFAC.txt:

 

; FACFORM Version No. 1.20,  May 4, 1992

; Run on 05-04-1992 12:44:12

; from Keyword file: TAPEkey.txt

Facets = 5

Title = Demonstration of Fixed Length data file

Output = tapeOUT.txt ; the output of the FACETS analysis

Models =       ; we model only subject-judge-item here

?,,?,,10,R     ;Item 10 has a rating (or partial credit) scale

?,,?,,?,D      ;Items 1-9 dichotomies

*

Labels =

1,Subject

1=Alice

2=Arthur

|

7=Mabel

8=Mary

*

2,Item Group

1=

2=

3=

4=

*

3,Judge

20=

21=

 

|

28=

29=

*

4,Session

1=

2=

3=

4=

*

5,Item

1=

2=

|

9=

10=

*

Data =

4,1,26,3,1-3,1,0,0

4,2,25,2,4-6,0,1,1

4,3,23,3,7-9,0,1,1

4,4,20,3,10,0

5,1,29,2,1-3,0,1,0

|

8,4,24,3,10,3

3,1,26,3,1-3,0,1,1

3,2,27,1,4-6,1,0,1

3,3,22,3,7-9,1,1,1

3,4,24,2,10,1