Re: st: option refseqid in command sqom

[Ulrich Kohler <kohler@wzb.eu>]

Am Mittwoch, den 29.09.2010, 08:06 -0700 schrieb zorobabel:
> Dear statalisters,
> I have a question about the option "refseqid" command sqom. In fact, i
> implemented this command but stata shows the following message error:
> 
> sqom,refseqid(0.76470846)
> option refseqid() not allowed
> r(198);


-sqom- is part of the user written ado package "sq" for sequence
analysis and optimal matching, as described in Stata Journal 6 (4). The
initial version distributed along with the article in the Stata Journal
has been updated on various occations via SSC. The most recent update
was in March 2010. Make sure that you use this most recent version, i.e.

. ssc install sq, replace

Although I think you use an outdated version of -sqom-, the update as
such will only give you a different error-message. From what you showed
us above I infer that you used a sequence identifier variable with
dezimal points, while -sqom- wants the sequence identifier to be an
integer number or a string variable: 

. use http://www.wzb.eu/~kohler/ado/youthemp, clear
. reshape long st, i(id) j(order)
. sqset st id order 
. sqom, refseqid(1)
. replace id = id/7
. sqom, refseqid(.2857143)
reference sequence does not exist

I going to look into the code of -sqom-, -sqset- and all the other
programs of the sq-package to see whether I can change the behaviour of
the program concerning non-integer values in sequence identifiers.
Meanwhile I suggest to use only integer sequence-id variables. In the
example above this could be done as follows:

. egen newid = group(id)
. sqset st newid order
. sqom, refseqid(1)


Man regards

Uli



-- 
kohler@wzb.eu
+49 (0)30 25491-361


*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/