Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
<<attachment: winmail.dat>>
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
st: Estimating cure models
From
Oras Alabas <[email protected]>
To
"'[email protected]'" <[email protected]>
Subject
st: Estimating cure models
Date
Thu, 14 Nov 2013 15:59:02 +0000
Hi
I am trying to fit a mixture cure fraction model in STATA for a cardiovascular coronary patients recruited between 2005 to 2010.
I have used the mixture cure fraction model by Paul Lambert using the following syntax: "strsmix, dist(weibull) link(identity) bhazard(rate), where rate is the probability of death from the background population.
Unfortunately, all models failed to converge showing the following error message "could not calculate numerical derivatives missing values encountered" although there is no missing values. By replacing the link function from identity to logistic the model run however the estimate of cure doesn't seems correct.
Could you please explain why the model fail to converge? Also can I ask for the STATA code of the cure fraction plot after strsmix?
Many Thanks
Oras
Dr Oras Alabas
Research Fellow in Epidemiology of Cardiovascular Disease
Leeds Institute of Genetics, Health and Therapeutics (LIGHT)
University of Leeds
Worsley Building | Room 8.49
Leeds, UK
LS2 9JT
Tel: 0113 343 8874
email: [email protected]