DATA LIST FILE Command
Locked
 
|
I have to read in some very large sets of data. Currently I am doing something like the following:
DATA LIST FILE = xyz RECORDS = 4
/resp = 1- 4
/
/mostoften 7 - 9
/current1 TO current31 7 - 99
...
And so on.
The text files that have the data are arranged by card. Columns 5 - 6 let me know the exact card for each respondent. So, in the above example, a 01, 02, 03, or 04 would appear in columns 5 - 6.
Is there a way in SPSS for me to read the card number and let DATA LIST FILE know to put the relevant card information in for each respondent? I greatly oversimplified the issue I am encountering here and wish to remove any thought that goes into the ordering of the subcommands above.
Thank you.
Zachary
|
|
|
Zachary: Does the card number change the record layout? I had a similar task using a text file—different record
length and variables depending on an identifier in each row. I read in all the data after the identifier as a long string. I used dataset copy to create versions of the data based on the
identifier, and SELECT IF to limit each dataset to one record layout. I then used NUMERIC and STRING to define the record layout, and a
series of commands like: COMPUTE var_nbr = NUMBER(SUBSTR(longstring,1,4),F8.0)
COMPUTE var_str =
SUBSTR(longstring,5,10) Does this help as an approach? It might be cumbersome if you
have a lot of variables to parse, but all the computes can be run on the same
pass. Jim Marks Director, Market Research x1616 From: SPSSX(r) Discussion
[mailto:[hidden email]] On Behalf Of Zachary Feinstein I have to read in some very large sets of data. Currently I am
doing something like the following: DATA LIST FILE = xyz RECORDS = 4 /resp = 1- 4 / /mostoften 7 - 9 /current1 TO current31 7 - 99 ... And so on. The text files that have the data are arranged by card.
Columns 5 - 6 let me know the exact card for each respondent. So, in the
above example, a 01, 02, 03, or 04 would appear in columns 5 - 6. Is there a way in SPSS for me to read the card number and let DATA
LIST FILE know to put the relevant card information in for each
respondent? I greatly oversimplified the issue I am encountering here and
wish to remove any thought that goes into the ordering of the subcommands
above. Thank you. Zachary |
|
|
In reply to this post by Zachary Feinstein
From your description, this sounds like a case for FILE TYPE MIXED. That allows you to specify different layouts for input records based on a record type id. Here is an example from the CSR. FILE TYPE MIXED FILE='/data/treatmnt.txt' RECORD=RECID 1-2. + RECORD TYPE 21,22,23,24. + DATA LIST /SEX 5 AGE 6-7 DOSAGE 8-10 RESULT 12. + RECORD TYPE 25. + DATA LIST /SEX 5 AGE 6-7 DOSAGE 10-12 RESULT 15. END FILE TYPE. If your scheme doesn't fit this model, you can always write an INPUT PROGRAM to build the cases HTH, Jon Peck SPSS, an IBM Company [hidden email] 312-651-3435
I have to read in some very large sets of data. Currently I am doing something like the following: DATA LIST FILE = xyz RECORDS = 4 /resp = 1- 4 / /mostoften 7 - 9 /current1 TO current31 7 - 99 ... And so on. The text files that have the data are arranged by card. Columns 5 - 6 let me know the exact card for each respondent. So, in the above example, a 01, 02, 03, or 04 would appear in columns 5 - 6. Is there a way in SPSS for me to read the card number and let DATA LIST FILE know to put the relevant card information in for each respondent? I greatly oversimplified the issue I am encountering here and wish to remove any thought that goes into the ordering of the subcommands above. Thank you. Zachary zsfeinstein@... |
Administrator
|
Here are a couple more examples using FILE TYPE MIXED.
www.angelfire.com/wv/bwhomedir/spss/read_data_from_mixed_file.txt Bruce
--
Bruce Weaver bweaver@lakeheadu.ca http://sites.google.com/a/lakeheadu.ca/bweaver/ "When all else fails, RTFM." PLEASE NOTE THE FOLLOWING: 1. My Hotmail account is not monitored regularly. To send me an e-mail, please use the address shown above. 2. The SPSSX Discussion forum on Nabble is no longer linked to the SPSSX-L listserv administered by UGA (https://listserv.uga.edu/). |
|
In reply to this post by Zachary Feinstein
Zachary
It's not exactly clear from your message what your
data actually look like, but if it's a straightforward rectangular matrix (ie
80-column Hollerith cards) there's a long step-by-step tutorial using DATA LIST
on:
(item 1.3.3) which takes you from raw data to SPSS
saved file. Line by line syntax with screenshots at each step of syntax
file and data editor.
This example is for one card per case (24
variables) but there's another tutorial using data from the British Social
Attitudes Survey (23 records per case) on:
This only extracts 2 variables, but the logic is
very clear and can be extended to your own data. I can send you the syntax
for all 900+ variables if you like!
It shouldn't really matter if your cards are of
different lengths. If you get stuck, send me an extract from your data so
I can better help you out.
John Hall
From: [hidden email]
|
| Free forum by Nabble | Edit this page |