Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

RE: st: Changing range in sts graph


From   "Allen Buxton" <ABUXTON@childrensoncologygroup.org>
To   <statalist@hsphsun2.harvard.edu>
Subject   RE: st: Changing range in sts graph
Date   Fri, 7 Dec 2007 09:12:04 -0800

I agree, messing with _t is a bad idea and I mess with it in my own
program (context) where _d & _t are preserved in the dataset. 
I will (re)visit tmax. 

-thanks.

Allen Buxton
CureSearch, Children's Oncology Group


-----Original Message-----
From: owner-statalist@hsphsun2.harvard.edu
[mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of Mark Lunt
Sent: Friday, December 07, 2007 4:19 AM
To: statalist@hsphsun2.harvard.edu
Subject: Re: st: Changing range in sts graph

Svend Juul wrote:
> Diego Bellavia <messadua@yahoo.it> wrote:
>
>   
>> I have a Kaplan Meyer curve I built with sts command. 
>> The curve is nice but I would like to reduce the range of time showed

>> to highlight the first part of the graph (where I have more events).
>> I tried with the xscale (range (0 100)) option but without success. 
>>     
>
> Maarten suggested:
>
>    sysuse cancer, clear
>    stset studytim,failure(died)
>    sts graph 
>    sts graph if _t < 20
>
> and Allen Buxton suggested:
>
>    local rangelimit=100
>    replace _d=0  if _t>`rangelimit'
>    replace _t=`rangelimit' if _t>`rangelimit'
>
> Nick got nervous:
>
> Doesn't any messing with _t mess up anything later with the same 
> survival dataset? Sounds a bad idea to me. See Maarten's solution.
>
> -----------------------------------------------------
>
> Allen's suggestion is right, Maarten's is wrong. It is easy to see 
> with -sts list- :
>
>    sysuse cancer, clear
>    stset studytim,failure(died)
>    sts list
>
>    // Maarten's suggestion:
>    sts list if  _t < 20    
>
>    // Allen's suggestion (simplified):
>    replace _d=0  if _t>20
>    replace _t=min(_t,20)
>    sts list
>
> In Maartens suggestion, persons surviving 20 months or more were 
> excluded from the analysis, leading to a far too pessimistic survival 
> estimate at 19 months. In Allen's suggestion they are correctly 
> censored at 20 months.
>
> Svend
>
>   
There is the option tmax, which truncates the display of the curve at a
given time, without needing to play with the data: should keep everyone
happy.

Mark

*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/


*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/



© Copyright 1996–2021 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index