Neil,
I don't have experience with -strel-, I'm afraid. Just from looking at
the -syntax- statement, the using option should not be specified with
parenthesis.
Try
strel br(0[0.25]3[.5]5) using data/stata/life_tables if(sex == 1), ///
mergeby(sex year age)
and this problem ought to disappear.
HTH,
Eva
2008/10/6 Neil Shephard <[email protected]>:
> Hi,
>
> This is a bit of a long shot, but I'm stumped at the moment and need to
> make some headway soon. If anyone has experience using -strel- any
> insights you can offer would be appreciated.
>
> I'm trying to use the -strel- command written by authors at the Cancer
> Survival Group at the London School of Hygeine a& Tropical Medicine (see
> http://www.lshtm.ac.uk/ncdeu/cancersurvival/tools/index.htm ). I'm
> slightly handicapped though as despite contacting the authors last I'm
> yet to hear back from them and am working with the ado-files left by a
> colleague who downloaded them. Unfortunately they only seem to have
> updated the most recent ado-file and not the associated help file so I
> don't have a help file for the version I'm using :-\
>
> To use the command data should be -stset- prior to running and the old
> help file specifically states...
>
> "Before using strel, the command stset is required to inform Stata of the
> key survival-time variables (see stset helpfile). Survival time must
> be declared indirectly by declaring age at entry or origin and age at
> exit. A survival time calculated by the user may be present in the dataset
> but must not be declared in stset (and therefore takes no part in the
> analysis)."
>
> Now assuming this hasn't changed between versions (and apologies if I've
> made an ass out of u and me) I have -stset- my data as per these
> instructions. I have the following variables which I -stset-
>
> . des death age_death age
>
> storage display value
> variable name type format label variable label
> ------------------------------------------------------------------------------------------------------------------------
> death byte %9.0g death Binary indicator showing
> who died
> age_death float %9.0g Age at Death (in years)
> age float %9.0g Age at Diagnosis (in years)
>
> . stset age_death, failure(death) origin(age)
>
> failure event: death != 0 & death < .
> obs. time interval: (origin, age_death]
> exit on or before: failure
> t for analysis: (time-origin)
> origin: time age
>
> ------------------------------------------------------------------------------
> 244729 total obs.
> 1 obs. end on or before enter()
> ------------------------------------------------------------------------------
> 244728 obs. remaining, representing
> 0 failures in single record/single failure data
> 608297.2 total analysis time at risk, at risk from t = 0
> earliest observed entry t = 0
> last observed exit t = 12.53388
>
> . st_show
>
> failure _d: death
> analysis time _t: (age_death-origin)
> origin: time age
>
> I have my life tables set-up in a separate file and am trying to run
> -strel- but can not get past the syntax....
>
> . set trace on
> . strel br(0[0.25]3[.5]5) using(data/stata/life_tables) if(sex == 1),
> mergeby(sex year age)
> --------------------------------------------------------------------------------------------------------
> begin strel ---
> - version 8.2
> - if replay() {
> syntax [, *]
> if _by() {
> error 190
> }
> if "`e(cmd2)'" != "strel" {
> error 301
> }
> ml display, `options'
> exit
> }
> - syntax anything(name=breaks id="numlist") using/ [if] [in] [iweight/],
> [ Mergeby(varlist) SHowrates at(numlist asc >=0
>> ) GRoup(integer 0) TRy(integer 3) Continue DIagdate(varname numeric)
> PERiod(numlist asc >1900 <2050 min=1 max=2 intege
>> r) HYbrid STANdardise(varlist min=1) STDprop(namelist min=1 max=1)
> BRenner SAving(string) replace level(integer 0) FIN
>> al DIVider SEParator(integer 0) noMODel * ]
> invalid 'data'
> ----------------------------------------------------------------------------------------------------------
> end strel ---
> r(198);
>
>
> I can't fathom where the "invalid 'data'" is coming from based on the
> syntax definintion? There doesn't appear to be any checks made on the
> data itself, simply that either breaks or a numlist are supplied along
> with the 'using lifetables'.
>
> Like I say, its a long shot but if anyone has experience of using
> -strel- and has any suggestions I'd be very grateful.
>
> Regards,
>
> Neil
>
>
> --
> "We should make things as simple as possible, but not simpler" - Anon (not Albert Einstein)
>
>
> ***********************************************************************
> This message may contain confidential and privileged information.
> If you are not the intended recipient you should not disclose, copy
> or distribute information in this e-mail or take any action in reliance
> on its contents. To do so is strictly prohibited and may be unlawful.
> Please inform the sender that this message has gone astray before
> deleting it. Thank you.
>
> 2008 marks the 60th anniversary of the NHS. It's an opportunity to pay
> tribute to the NHS staff and volunteers who help shape the service, and
> celebrate their achievements.
>
> If you work for the NHS and would like an NHSmail email account, go
> to: www.connectingforhealth.nhs.uk/nhsmail
> ***********************************************************************
>
> *
> * 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/
>
*
* 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/
