OVERALS Nonlinear Canonical Correlation Error Message

classic Classic list List threaded Threaded
2 messages Options
Reply | Threaded
Open this post in threaded view
|

OVERALS Nonlinear Canonical Correlation Error Message

Glenys Lafrance
Hi Listers,

I am trying to compute a canonical correlation using the OVERALS command.
I get the following warning:

"Command name: OVERALS The object(s) listed above must be removed since
they contain missing (or out of range) values in all sets.  This command
not executed." Note:  there is no info ABOVE this warning.

The case processing box indicates that a number of cases will be dropped
(I presume listwise deletion is built-in to the procedure)...but I know I
have at least 180 valid cases left.  Here is a copy of the syntax log in
case someone can tell if there's a problem specifying options:

 OVERALS
  /VARIABLES=q11a(4) q11b(4) q11c(4) q11d(4) q11e(4) q12a(4) q12b(4) q12c
(4) q12d(4) q12e(4) q12f(4) q12g(4)
  /ANALYSIS=q11a(NUME) q11b(NUME) q11c(NUME) q11d(NUME) q11e(NUME) q12a
(NUME) q12b(NUME) q12c(NUME) q12d(NUME) q12e(NUME)
  q12f(NUME) q12g(NUME)
  /SETS 2 (5, 7)
  /DIMENSION=3
  /PRINT FREQ CENTROID HISTORY WEIGHTS FIT QUANT OBJECT
  /PLOT QUANT OBJECT LOADINGS CENTROID( q11a q11b q11c q11d q11e q12a q12b
q12c q12d q12e q12f q12g )  TRANS NDIM(1,MAX)
  /MAXITER = 100
  /CONVERGENCE = .00001
  /INITIAL=RANDOM .

My data is actually a 4-point ordinal scale..so I don't think I should use
the Manova version of cca.  Any help is appreciated.

Glenys Lafrance
Reply | Threaded
Open this post in threaded view
|

Re: OVERALS Nonlinear Canonical Correlation Error Message

Kooij, A.J. van der
Hi Glenys,
 
The note at the case processing table says that the listed cases were removed. They are, but the warning acts as an error, causing the program to stop execution, which is a bug, that we will fix asap.
So, for now you have to remove the listed cases yourself. In the warning "above" should be "listed below case processing table ".
 
Regards,
Anita van der Kooij
Data Theory Group
Leiden University

________________________________

From: SPSSX(r) Discussion on behalf of Glenys Lafrance
Sent: Sun 26/11/2006 11:39
To: [hidden email]
Subject: OVERALS Nonlinear Canonical Correlation Error Message



Hi Listers,

I am trying to compute a canonical correlation using the OVERALS command.
I get the following warning:

"Command name: OVERALS The object(s) listed above must be removed since
they contain missing (or out of range) values in all sets.  This command
not executed." Note:  there is no info ABOVE this warning.

The case processing box indicates that a number of cases will be dropped
(I presume listwise deletion is built-in to the procedure)...but I know I
have at least 180 valid cases left.  Here is a copy of the syntax log in
case someone can tell if there's a problem specifying options:

 OVERALS
  /VARIABLES=q11a(4) q11b(4) q11c(4) q11d(4) q11e(4) q12a(4) q12b(4) q12c
(4) q12d(4) q12e(4) q12f(4) q12g(4)
  /ANALYSIS=q11a(NUME) q11b(NUME) q11c(NUME) q11d(NUME) q11e(NUME) q12a
(NUME) q12b(NUME) q12c(NUME) q12d(NUME) q12e(NUME)
  q12f(NUME) q12g(NUME)
  /SETS 2 (5, 7)
  /DIMENSION=3
  /PRINT FREQ CENTROID HISTORY WEIGHTS FIT QUANT OBJECT
  /PLOT QUANT OBJECT LOADINGS CENTROID( q11a q11b q11c q11d q11e q12a q12b
q12c q12d q12e q12f q12g )  TRANS NDIM(1,MAX)
  /MAXITER = 100
  /CONVERGENCE = .00001
  /INITIAL=RANDOM .

My data is actually a 4-point ordinal scale..so I don't think I should use
the Manova version of cca.  Any help is appreciated.

Glenys Lafrance



**********************************************************************
This email and any files transmitted with it are confidential and
intended solely for the use of the individual or entity to whom they
are addressed. If you have received this email in error please notify
the system manager.
**********************************************************************