estimation of lambda and gamma with std errors for a weibull model
On Wed, Jan 14, 2004 at 10:45:56PM +0100, G?ran Brostr?m wrote:
On Wed, Jan 14, 2004 at 09:10:51PM +0100, Fredrik Lundgren wrote:
Dear R experts, How should lambda and gamma (with std.errors) be calculated for a weibull model with age as an independent predictor? I have assumed that this can be done with survreg with e. g. (summary(survreg(Surv(time, status) ~ age, dist = 'weibull')) ) and predict.survreg with e.g. (predict(model, se.fit = T, newdata = data.frame(age = seq(50, 80, 5)) but unfortunately I'm uncapable to sort out how to get the lambda and gamma values (with std.errors). I haven't found any example of this in the help pages and would really appreciate any help!
In my package 'eha', function 'weibreg', you will find short discussion of the different parametrizations of the Weibull distribution. Weibull (in base) and weibreg (eha) use the same parametrization, different from the one in survreg. See the help page for weibreg. Oops, I can spot an error in that page; the reference to 'dgamma' should really be to 'dweibull'. G?ran
To elaborate further, you should maybe be satisfied with the standard errors you get on the log scale. Calculate confidence intervals (or whatever you want the se's for) on that scale, and transform these intervals to any scale you like. Usually much better than doing it in reverse order, ie, calculating se's (via the delta method) for 'lambda' and 'gamma', and then the confidence intervals. G?ran
With best wishes and thanks in advance for any help Fredrik Lundgren
______________________________________________ R-help at stat.math.ethz.ch mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
-- G?ran Brostr?m tel: +46 90 786 5223 Department of Statistics fax: +46 90 786 6614 Ume? University http://www.stat.umu.se/egna/gb/ SE-90187 Ume?, Sweden e-mail: gb at stat.umu.se
______________________________________________ R-help at stat.math.ethz.ch mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
G?ran Brostr?m tel: +46 90 786 5223 Department of Statistics fax: +46 90 786 6614 Ume? University http://www.stat.umu.se/egna/gb/ SE-90187 Ume?, Sweden e-mail: gb at stat.umu.se