CTABLES Problem
Locked
 
|
Can anyone tell me why the syntax below is producing the table for only one of the values of SUBJECTID?
I am using BY BY SUBJECTID in lieu of SPLIT FILE BY SUBJECTID which is not supported.
CTABLES /VLABELS VARIABLES=R1_1 R2_1 R3_1 R4_1 R5_1 R6_1 R7_1 R8_1 R9_1 R10_1 R11_1 R12_1 R13_1 R14_1 R15_1 R17_1 R18_1 R19_1 R20_1 R21_1 DISPLAY=LABEL /TABLE (R1_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R2_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R3_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R4_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R5_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R6_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R7_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R8_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R9_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R10_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R11_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R12_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R13_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R14_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R15_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R17_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R18_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R19_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R20_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R21_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0]) BY BY SUBJECTID /SLABELS POSITION=ROW VISIBLE=YES /CLABELS ROWLABELS=OPPOSITE /CATEGORIES VARIABLES=SUBJECTID R1_1 R2_1 R3_1 R4_1 R5_1 R6_1 R7_1 R8_1 R9_1 R10_1 R11_1 R12_1 R13_1 R14_1 R15_1 R17_1 R18_1 R19_1 R20_1 R21_1 ORDER=A KEY=LABEL MISSING=EXCLUDE EMPTY=EXCLUDE. Prof Peter Cuttance Director Research Australia Development and Innovation Institute (radii.org) Ph. +613 9534 2934 Mob. +61 428 990 904 Fax. +613 8640 0741 Suite 71 197 Canterbury Road St Kilda West Victoria AUSTRALIA 3182 |
|
|
Peter,
If you haven't already received useful replies, would you please say what you are wanting the table to look like. I have used ctables only a few times so there is a lot I don't know, but in looking at the syntax reference (v18, p. 490) the BY BY specification creates layers. Could it be that you have what you want but don't realize it? Gene Maguin >>Can anyone tell me why the syntax below is producing the table for only one of the values of SUBJECTID? I am using BY BY SUBJECTID in lieu of SPLIT FILE BY SUBJECTID which is not supported. CTABLES /VLABELS VARIABLES=R1_1 R2_1 R3_1 R4_1 R5_1 R6_1 R7_1 R8_1 R9_1 R10_1 R11_1 R12_1 R13_1 R14_1 R15_1 R17_1 R18_1 R19_1 R20_1 R21_1 DISPLAY=LABEL /TABLE (R1_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R2_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R3_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R4_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R5_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R6_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R7_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R8_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R9_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R10_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R11_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R12_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R13_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R14_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R15_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R17_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R18_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R19_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R20_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R21_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0]) BY BY SUBJECTID /SLABELS POSITION=ROW VISIBLE=YES /CLABELS ROWLABELS=OPPOSITE /CATEGORIES VARIABLES=SUBJECTID R1_1 R2_1 R3_1 R4_1 R5_1 R6_1 R7_1 R8_1 R9_1 R10_1 R11_1 R12_1 R13_1 R14_1 R15_1 R17_1 R18_1 R19_1 R20_1 R21_1 ORDER=A KEY=LABEL MISSING=EXCLUDE EMPTY=EXCLUDE. Prof Peter Cuttance Director Research Australia Development and Innovation Institute (radii.org) Ph. +613 9534 2934 Mob. +61 428 990 904 Fax. +613 8640 0741 Suite 71 197 Canterbury Road St Kilda West Victoria AUSTRALIA 3182 ===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD |
|
|
In reply to this post by Peter Cuttance
Sounds like you really want one table at a
time then I would use something like: TEMPORARY . SELECT IF (SubjectID= ) . (table syntax here) Or use FILTER rather then the SELECT IF You can use DO REPEAT logic if you have many tables to generate: Or can embed the CTABLES command in a macro DO REPEAT s = subject1 subject2 . . . . + TEMPORARY . + SELECT IF (SubjectID=s) . + (table command here) END REPEAT . Note that SELECT IF and FILTER do not require data to be presorted as
SPLIT FILES does. I rarely use SPLIT FILES personally as my break vars are often not
usually mutually exclusive Dennis From: Peter Cuttance
[mailto:[hidden email]] Hi Dennis SUBJECTID is the ID for the subject the student is studying —
each student has responded to a standard survey for each subject they have
studied. Dropping BY BY and replacing it with BY simply results in a table that
is N tables wide (where N is the number of separate SUBJECTID's. I want a separate set of table results for the survey for each
SUBJECTID — nornally this would be achieved by SPLIT FILE BY SUBJECTID, but
this is not supported in CTABLES. Any suggestions would be welcome. Regards Peter Research Australia
Development and Innovation Institute (radii.org) Ph. +613
9534 2934 Mob. +61 428 990 904 Fax. +613 8640
0741 St Kilda West
On 28/01/2010, at 4:59 PM, Dennis Deck wrote: Not sure why you are trying “BY
BY” — a single “BY” is the proper syntax. Perhaps you have a good reason to break
the summary by SUBJECTID but it seems suspect. Do you have more than one case per
subject? And why would you average by subject anyway? I would expect a variable like Group or
StudyCondition or Time . . .. A simple command for checking results, of
course, would be: CROSSTABS Tables=
R1_1 R2_1 R3_1 R4_1 R5_1
R6_1 R7_1 R8_1 R9_1 R10_1 R11_1 R12_1
R13_1 R14_1 R15_1 R17_1 R18_1 R19_1 R20_1 R21_1 by Group /Cells= Count Row . Dennis Deck RMC Research From: Peter Cuttance [[hidden email]] Can anyone tell me why the syntax below is
producing the table for only one of the values of SUBJECTID? I am using BY BY SUBJECTID in lieu of
SPLIT FILE BY SUBJECTID which is not supported. CTABLES /VLABELS VARIABLES=R1_1 R2_1
R3_1 R4_1 R5_1 R6_1 R7_1 R8_1 R9_1 R10_1 R11_1 R12_1 R13_1 R14_1 R15_1 R17_1
R18_1 R19_1 R20_1 R21_1 DISPLAY=LABEL /TABLE (R1_1 [COUNT F40.0,
ROWPCT.COUNT PCT40.0] + R2_1 [COUNT F40.0,
ROWPCT.COUNT PCT40.0] + R3_1 [COUNT F40.0,
ROWPCT.COUNT PCT40.0] + R4_1 [COUNT F40.0,
ROWPCT.COUNT PCT40.0] + R5_1 [COUNT F40.0,
ROWPCT.COUNT PCT40.0] + R6_1 [COUNT F40.0,
ROWPCT.COUNT PCT40.0] + R7_1 [COUNT F40.0,
ROWPCT.COUNT PCT40.0] + R8_1 [COUNT F40.0,
ROWPCT.COUNT PCT40.0] + R9_1 [COUNT F40.0,
ROWPCT.COUNT PCT40.0] + R10_1 [COUNT F40.0,
ROWPCT.COUNT PCT40.0] + R11_1 [COUNT F40.0,
ROWPCT.COUNT PCT40.0] + R12_1 [COUNT F40.0,
ROWPCT.COUNT PCT40.0] + R13_1 [COUNT F40.0,
ROWPCT.COUNT PCT40.0] + R14_1 [COUNT F40.0,
ROWPCT.COUNT PCT40.0] + R15_1 [COUNT F40.0,
ROWPCT.COUNT PCT40.0] + R17_1 [COUNT F40.0,
ROWPCT.COUNT PCT40.0] + R18_1 [COUNT F40.0,
ROWPCT.COUNT PCT40.0] + R19_1 [COUNT F40.0,
ROWPCT.COUNT PCT40.0] + R20_1 [COUNT F40.0,
ROWPCT.COUNT PCT40.0] + R21_1 [COUNT F40.0,
ROWPCT.COUNT PCT40.0]) BY BY SUBJECTID /SLABELS POSITION=ROW
VISIBLE=YES /CLABELS ROWLABELS=OPPOSITE /CATEGORIES
VARIABLES=SUBJECTID R1_1 R2_1 R3_1 R4_1 R5_1 R6_1 R7_1 R8_1 R9_1 R10_1 R11_1
R12_1 R13_1 R14_1 R15_1 R17_1 R18_1 R19_1 R20_1 R21_1 ORDER=A KEY=LABEL
MISSING=EXCLUDE EMPTY=EXCLUDE. |
|
|
In reply to this post by Peter Cuttance
If you have a list of the subjects (SUBJECTID)
then simply copy that into the DO REPEAT logic. DO REPEAT s = subject1 subject2
. . . . Takes all of 2 minutes You might be able to get CTABLES to place
the next subject below rather than to the right. One thought was to add a dummy variable to
force CTABLES to put the next subject below rather to the right (not sure if that will work but easy to
try) COMPUTE Dummy = 1 . CTABLES . . . .. BY Dummy BY SubjectID
. Dennis From: Peter Cuttance
[mailto:[hidden email]] Hi Dennis That would work OK — but, there are 300+ tables, so I am looking
for something less laborious. Cheers Peter Prof Peter Cuttance Director Research Australia
Development and Innovation Institute (radii.org) Ph. +613
9534 2934 Mob. +61 428 990 904 Fax. +613 8640
0741 St Kilda West
On 29/01/2010, at 4:45 AM, Dennis Deck wrote:
Sounds like you really want one table at a
time then I would use something like: TEMPORARY . SELECT IF (SubjectID= ) . (table syntax here) Or use FILTER rather then the SELECT IF You can use DO REPEAT logic if you have
many tables to generate: Or can embed the CTABLES command in a macro DO REPEAT s = subject1 subject2
. . . . + TEMPORARY . + SELECT IF (SubjectID=s) . + (table command here) END REPEAT . Note that SELECT IF and FILTER do not
require data to be presorted as SPLIT FILES does. I rarely use SPLIT FILES personally as my
break vars are often not usually mutually exclusive Dennis From: Peter Cuttance [[hidden email]] Hi Dennis SUBJECTID is the ID for the subject the
student is studying — each student has responded to a standard survey for
each subject they have studied. Dropping BY BY and replacing it with BY
simply results in a table that is N tables wide (where N is the number of
separate SUBJECTID's. I want a separate set of table results for
the survey for each SUBJECTID — nornally this would be achieved by SPLIT
FILE BY SUBJECTID, but this is not supported in CTABLES. Any suggestions would be welcome. Regards Peter Research Australia
Development and Innovation Institute (radii.org) Ph. +613
9534 2934 Mob. +61 428 990 904 Fax. +613 8640
0741 St Kilda West
On 28/01/2010, at 4:59 PM, Dennis Deck
wrote:
Not sure why you are trying “BY
BY” — a single “BY” is the proper syntax. Perhaps you have a good reason to break
the summary by SUBJECTID but it seems suspect. Do you have more than one case per
subject? And why would you average by subject anyway? I would expect a variable like Group or
StudyCondition or Time . . .. A simple command for checking results, of
course, would be: CROSSTABS Tables=
R1_1 R2_1 R3_1 R4_1 R5_1
R6_1 R7_1 R8_1 R9_1 R10_1 R11_1 R12_1
R13_1 R14_1 R15_1 R17_1 R18_1 R19_1 R20_1 R21_1 by Group /Cells= Count Row . Dennis Deck RMC Research From: Peter Cuttance [[hidden email]] Can anyone tell me why the syntax below is
producing the table for only one of the values of SUBJECTID? I am using BY BY SUBJECTID in lieu of
SPLIT FILE BY SUBJECTID which is not supported. CTABLES /VLABELS VARIABLES=R1_1 R2_1
R3_1 R4_1 R5_1 R6_1 R7_1 R8_1 R9_1 R10_1 R11_1 R12_1 R13_1 R14_1 R15_1 R17_1
R18_1 R19_1 R20_1 R21_1 DISPLAY=LABEL /TABLE (R1_1 [COUNT F40.0,
ROWPCT.COUNT PCT40.0] + R2_1 [COUNT F40.0,
ROWPCT.COUNT PCT40.0] + R3_1 [COUNT F40.0,
ROWPCT.COUNT PCT40.0] + R4_1 [COUNT F40.0,
ROWPCT.COUNT PCT40.0] + R5_1 [COUNT F40.0,
ROWPCT.COUNT PCT40.0] + R6_1 [COUNT F40.0,
ROWPCT.COUNT PCT40.0] + R7_1 [COUNT F40.0,
ROWPCT.COUNT PCT40.0] + R8_1 [COUNT F40.0,
ROWPCT.COUNT PCT40.0] + R9_1 [COUNT F40.0, ROWPCT.COUNT
PCT40.0] + R10_1 [COUNT F40.0,
ROWPCT.COUNT PCT40.0] + R11_1 [COUNT F40.0,
ROWPCT.COUNT PCT40.0] + R12_1 [COUNT F40.0,
ROWPCT.COUNT PCT40.0] + R13_1 [COUNT F40.0,
ROWPCT.COUNT PCT40.0] + R14_1 [COUNT F40.0,
ROWPCT.COUNT PCT40.0] + R15_1 [COUNT F40.0,
ROWPCT.COUNT PCT40.0] + R17_1 [COUNT F40.0,
ROWPCT.COUNT PCT40.0] + R18_1 [COUNT F40.0,
ROWPCT.COUNT PCT40.0] + R19_1 [COUNT F40.0,
ROWPCT.COUNT PCT40.0] + R20_1 [COUNT F40.0,
ROWPCT.COUNT PCT40.0] + R21_1 [COUNT F40.0,
ROWPCT.COUNT PCT40.0]) BY BY SUBJECTID /SLABELS POSITION=ROW
VISIBLE=YES /CLABELS ROWLABELS=OPPOSITE /CATEGORIES
VARIABLES=SUBJECTID R1_1 R2_1 R3_1 R4_1 R5_1 R6_1 R7_1 R8_1 R9_1 R10_1 R11_1
R12_1 R13_1 R14_1 R15_1 R17_1 R18_1 R19_1 R20_1 R21_1 ORDER=A KEY=LABEL
MISSING=EXCLUDE EMPTY=EXCLUDE. |
|
|
If you don't mind doing a data pass for each subjectid, there is a simple solution using programmability. Here's an example using the cars.sav file. begin program. import spss, spssaux2 cmd="""CTABLES /VLABELS VARIABLES=year origin DISPLAY=DEFAULT /TABLE year [COUNT F40.0] by cylinder""" spssaux2.generalizedSplit("origin", cmd, vartype=0) end program. cmd is whatever syntax you want. generalizedSplit is a function in the spssaux2.py module available from SPSS Developer Central. In this case, I'm splitting on the origin variable. vartype=0 means that the split variable is numeric. Data do not need to be sorted. Regards, Jon Peck SPSS, an IBM Company [hidden email] 312-651-3435
If you have a list of the subjects (SUBJECTID) then simply copy that into the DO REPEAT logic. DO REPEAT s = subject1 subject2 . . . . Takes all of 2 minutes You might be able to get CTABLES to place the next subject below rather than to the right. One thought was to add a dummy variable to force CTABLES to put the next subject below rather to the right (not sure if that will work but easy to try) COMPUTE Dummy = 1 . CTABLES . . . .. BY Dummy BY SubjectID . Dennis From: Peter Cuttance [mailto:cuttance@...] Sent: Thursday, January 28, 2010 11:08 AM To: Dennis Deck Subject: Re: CTABLES Problem Hi Dennis That would work OK — but, there are 300+ tables, so I am looking for something less laborious. Cheers Peter Prof Peter Cuttance Director Research Australia Development and Innovation Institute (radii.org) Ph. +613 9534 2934 Mob. +61 428 990 904 Fax. +613 8640 0741 Suite 71 197 Canterbury Road St Kilda West Victoria AUSTRALIA 3182 On 29/01/2010, at 4:45 AM, Dennis Deck wrote: Sounds like you really want one table at a time then I would use something like: TEMPORARY . SELECT IF (SubjectID= ) . (table syntax here) Or use FILTER rather then the SELECT IF You can use DO REPEAT logic if you have many tables to generate: Or can embed the CTABLES command in a macro DO REPEAT s = subject1 subject2 . . . . + TEMPORARY . + SELECT IF (SubjectID=s) . + (table command here) END REPEAT . Note that SELECT IF and FILTER do not require data to be presorted as SPLIT FILES does. I rarely use SPLIT FILES personally as my break vars are often not usually mutually exclusive Dennis From: Peter Cuttance [mailto:cuttance@...] Sent: Thursday, January 28, 2010 3:44 AM To: Dennis Deck Subject: Re: CTABLES Problem Hi Dennis SUBJECTID is the ID for the subject the student is studying — each student has responded to a standard survey for each subject they have studied. Dropping BY BY and replacing it with BY simply results in a table that is N tables wide (where N is the number of separate SUBJECTID's. I want a separate set of table results for the survey for each SUBJECTID — nornally this would be achieved by SPLIT FILE BY SUBJECTID, but this is not supported in CTABLES. Any suggestions would be welcome. Regards Peter Research Australia Development and Innovation Institute (radii.org) Ph. +613 9534 2934 Mob. +61 428 990 904 Fax. +613 8640 0741 Suite 71 197 Canterbury Road St Kilda West Victoria AUSTRALIA 3182 On 28/01/2010, at 4:59 PM, Dennis Deck wrote: Not sure why you are trying “BY BY” — a single “BY” is the proper syntax. Perhaps you have a good reason to break the summary by SUBJECTID but it seems suspect. Do you have more than one case per subject? And why would you average by subject anyway? I would expect a variable like Group or StudyCondition or Time . . .. A simple command for checking results, of course, would be: CROSSTABS Tables= R1_1 R2_1 R3_1 R4_1 R5_1 R6_1 R7_1 R8_1 R9_1 R10_1 R11_1 R12_1 R13_1 R14_1 R15_1 R17_1 R18_1 R19_1 R20_1 R21_1 by Group /Cells= Count Row . Dennis Deck RMC Research From: Peter Cuttance [mailto:cuttance@...] Sent: Wednesday, January 27, 2010 12:48 PM Subject: CTABLES Problem Can anyone tell me why the syntax below is producing the table for only one of the values of SUBJECTID? I am using BY BY SUBJECTID in lieu of SPLIT FILE BY SUBJECTID which is not supported. CTABLES /VLABELS VARIABLES=R1_1 R2_1 R3_1 R4_1 R5_1 R6_1 R7_1 R8_1 R9_1 R10_1 R11_1 R12_1 R13_1 R14_1 R15_1 R17_1 R18_1 R19_1 R20_1 R21_1 DISPLAY=LABEL /TABLE (R1_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R2_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R3_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R4_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R5_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R6_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R7_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R8_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R9_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R10_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R11_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R12_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R13_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R14_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R15_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R17_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R18_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R19_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R20_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R21_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0]) BY BY SUBJECTID /SLABELS POSITION=ROW VISIBLE=YES /CLABELS ROWLABELS=OPPOSITE /CATEGORIES VARIABLES=SUBJECTID R1_1 R2_1 R3_1 R4_1 R5_1 R6_1 R7_1 R8_1 R9_1 R10_1 R11_1 R12_1 R13_1 R14_1 R15_1 R17_1 R18_1 R19_1 R20_1 R21_1 ORDER=A KEY=LABEL MISSING=EXCLUDE EMPTY=EXCLUDE. |
|
In reply to this post by Dennis Deck
Dear Dennis,
DO REP is not going to work here. To the best of my knowledge you can only use it for data transformations, not for generating any output. You could use a very simple macro in order to accomplish this task. HTH and have a shiny happy weekend ;-D *Generate fake data. inp pro. loop id=1 to 30. end cas. end loop. end fil. end inp pro. ran id /nti (3). ren var nid=group. comp v1=rv.ber(.4). comp v2=rv.ber(.6). exe. *Do rep won't work but will delete all data. *do rep a=1 to 3. *temp. *sel if group=a. *cro v1 by v2. *end rep. *exe. *Very ugly macro definition. define !cross() !do !i = 1 !to 3 temp. sel if group=!i. cro v1 by v2. !doend !end. *Macro call will now generate desired output. !cross. Ruben van den Berg Methodologist TNS NIPO P: +31 20 522 5738 Date: Thu, 28 Jan 2010 18:08:50 +0000 From: [hidden email] Subject: FW: CTABLES Problem To: [hidden email] Sounds like you really want one table at a time then I would use something like:
TEMPORARY . SELECT IF (SubjectID= ) . (table syntax here)
Or use FILTER rather then the SELECT IF
You can use DO REPEAT logic if you have many tables to generate: Or can embed the CTABLES command in a macro
DO REPEAT s = subject1 subject2 . . . . + TEMPORARY . + SELECT IF (SubjectID=s) . + (table command here) END REPEAT .
Note that SELECT IF and FILTER do not require data to be presorted as SPLIT FILES does. I rarely use SPLIT FILES personally as my break vars are often not usually mutually exclusive
Dennis From: Peter Cuttance [mailto:[hidden email]]
Hi Dennis
SUBJECTID is the ID for the subject the student is studying — each student has responded to a standard survey for each subject they have studied.
Dropping BY BY and replacing it with BY simply results in a table that is N tables wide (where N is the number of separate SUBJECTID's.
I want a separate set of table results for the survey for each SUBJECTID — nornally this would be achieved by SPLIT FILE BY SUBJECTID, but this is not supported in CTABLES.
Any suggestions would be welcome.
Regards Peter
Research Australia Development and Innovation Institute (radii.org) Ph. +613 9534 2934 Mob. +61 428 990 904 Fax. +613 8640 0741 Suite 71 197 Canterbury Road St Kilda West Victoria AUSTRALIA 3182
On 28/01/2010, at 4:59 PM, Dennis Deck wrote:
Not sure why you are trying “BY BY” — a single “BY” is the proper syntax.
Perhaps you have a good reason to break the summary by SUBJECTID but it seems suspect. Do you have more than one case per subject? And why would you average by subject anyway? I would expect a variable like Group or StudyCondition or Time . . ..
A simple command for checking results, of course, would be:
CROSSTABS Tables= R1_1 R2_1 R3_1 R4_1 R5_1 R6_1 R7_1 R8_1 R9_1 R10_1 R11_1 R12_1 R13_1 R14_1 R15_1 R17_1 R18_1 R19_1 R20_1 R21_1 by Group /Cells= Count Row .
Dennis Deck RMC Research
From: Peter Cuttance [[hidden email]]
Can anyone tell me why the syntax below is producing the table for only one of the values of SUBJECTID?
I am using BY BY SUBJECTID in lieu of SPLIT FILE BY SUBJECTID which is not supported.
CTABLES /VLABELS VARIABLES=R1_1 R2_1 R3_1 R4_1 R5_1 R6_1 R7_1 R8_1 R9_1 R10_1 R11_1 R12_1 R13_1 R14_1 R15_1 R17_1 R18_1 R19_1 R20_1 R21_1 DISPLAY=LABEL /TABLE (R1_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R2_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R3_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R4_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R5_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R6_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R7_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R8_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R9_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R10_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R11_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R12_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R13_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R14_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R15_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R17_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R18_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R19_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R20_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0] + R21_1 [COUNT F40.0, ROWPCT.COUNT PCT40.0]) BY BY SUBJECTID /SLABELS POSITION=ROW VISIBLE=YES /CLABELS ROWLABELS=OPPOSITE /CATEGORIES VARIABLES=SUBJECTID R1_1 R2_1 R3_1 R4_1 R5_1 R6_1 R7_1 R8_1 R9_1 R10_1 R11_1 R12_1 R13_1 R14_1 R15_1 R17_1 R18_1 R19_1 R20_1 R21_1 ORDER=A KEY=LABEL MISSING=EXCLUDE EMPTY=EXCLUDE.
Express yourself instantly with MSN Messenger! MSN Messenger |
Administrator
|
Or alternatively... sort cases by group. split file by group. crosstabs v1 by v2. split file off. I assume CTABLES works with SPLIT FILE, does it not?
--
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/). |
| Free forum by Nabble | Edit this page |