see if this untested syntax does what you want. (be sure to check
whether there are off diagonal 1.00s.
dataset activate dscorrs.
missing data all (1.00).
That way you can avoid misuse of sysmis and procedures will still
process those values as missing.
Just curious what procedures are you doing that cannot deal with 1.00 on
the diagonal?
Art Kendall
Social Research Consultants
On 12/7/2010 2:42 PM, J Sutar wrote:
> Hi,
> I'm trying to capture a correlation matrix using OMS and then, for
> simplicity say, recode the ones in diagonal as sysmiss. However to do
> this I must know the variable names OMS decides to name them. As far
> as I am aware there is no option to allow any particular preference of
> the users choice? As it currently stands it concatenates the variable
> name and variable label. However this is extremely difficult to code
> within to a macro so that I can specify the variables on my RECODE
> command.
> sample code:
> get file ='C:\Program Files (x86)\SPSS\employee data.sav'.
> dataset name MyActiveDS.
>
> preserve.
> set tnumbers=both tvars=both.
> dataset declare dsCorrs.
> oms
> /select tables
> /if commands = ["correlations"]
> subtypes = ["correlations"]
> /destination format = sav numbered = tablenumber_
> outfile = dsCorrs /tag="dsCorrs".
> correl bdate educ jobcat salary salbegin jobtime prevexp minority.
> omsend tag=["dsCorrs"].
> restore.
> Setting the variable labels off (set tvars=name) is not an option as
> it would remove the variable labels from the rows also, which is no good.
> Any suggestion?
> Many thanks in advance.
> Jignesh
=====================
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
Art Kendall
Social Research Consultants
Locked
