Jari Haukka
2007-Mar-20 10:13 UTC
[R] Any R function for self-controlled case series method /effect absorption?
Hello, Has anyone written R functions for applying self-controlled case series methods (http://statistics.open.ac.uk/sccs/). In fact only thing needed is to modify glm function to allow absorption of effect. Eg. in Poisson model individual effect is used as factor, but it is considered as nuisance term where parameter estimates are not needed. Could anyone point how absorbing individual effect could be carried out in glm. There is already code for Stata (http://statistics.open.ac.uk/sccs/stata/aglm.ado), Genstat (http://statistics.open.ac.uk/sccs/genstat%5Csccs.gen), Glim (http://statistics.open.ac.uk/sccs/glim/SCCS.MAC) , and SAS (http://statistics.open.ac.uk/sccs/sas/macro/poisreg.sas). Regards, Jari Haukka
vito muggeo
2007-Mar-20 10:43 UTC
[R] Any R function for self-controlled case series method /effect absorption?
Dear Jari The problem is to build the dataset to apply the conditional logit model. However, as far as I know, no R function exists. BTW if you are dealing with time series of pollution and health, the following two papers might be of interest of you: It appears that the time series approach could be preferred. Heather J. Whitaker, Mounia N. Hocine, C. Paddy Farrington On case-crossover methods for environmental time series data Environmetrics Volume 18, Issue 2, Date: March 2007, Pages: 157-171 Lu, Zeger. On the equivalence of case-crossover and time series methods in environmental epidemiology Biostatistics, Early view Jari Haukka wrote:> Hello, > > Has anyone written R functions for applying self-controlled case series > methods (http://statistics.open.ac.uk/sccs/). > > In fact only thing needed is to modify glm function to allow absorption > of effect. Eg. in Poisson model individual effect is used as factor, but > it is considered as nuisance term where parameter estimates are not needed. > > Could anyone point how absorbing individual effect could be carried out > in glm. > > There is already code for Stata > (http://statistics.open.ac.uk/sccs/stata/aglm.ado), Genstat > (http://statistics.open.ac.uk/sccs/genstat%5Csccs.gen), Glim > (http://statistics.open.ac.uk/sccs/glim/SCCS.MAC) , and SAS > (http://statistics.open.ac.uk/sccs/sas/macro/poisreg.sas). > > > Regards, > Jari Haukka > > ______________________________________________ > R-help at stat.math.ethz.ch mailing list > https://stat.ethz.ch/mailman/listinfo/r-help > PLEASE do read the posting guide http://www.R-project.org/posting-guide.html > and provide commented, minimal, self-contained, reproducible code. > >-- ===================================Vito M.R. Muggeo Dip.to Sc Statist e Matem `Vianelli' Universit? di Palermo viale delle Scienze, edificio 13 90128 Palermo - ITALY tel: 091 6626240 fax: 091 485726/485612
Jari Haukka
2007-Mar-20 11:58 UTC
[R] Any R function for self-controlled case series method /effect
Dear Vito, I found found package "gnm" with function "gnm". The "eliminate" argument does exactly what I needed. Conditional Poisson models are handy to fit by gnm. Regards, Jari vito muggeo wrote:> Dear Jari > The problem is to build the dataset to apply the conditional logit > model. However, as far as I know, no R function exists. > > BTW if you are dealing with time series of pollution and health, the > following two papers might be of interest of you: > It appears that the time series approach could be preferred. > > > Heather J. Whitaker, Mounia N. Hocine, C. Paddy Farrington > On case-crossover methods for environmental time series data > Environmetrics > Volume 18, Issue 2, Date: March 2007, Pages: 157-171 > > Lu, Zeger. On the equivalence of case-crossover and time series > methods in environmental epidemiology Biostatistics, Early view > > > Jari Haukka wrote: >> Hello, >> >> Has anyone written R functions for applying self-controlled case >> series methods (http://statistics.open.ac.uk/sccs/). >> >> In fact only thing needed is to modify glm function to allow >> absorption of effect. Eg. in Poisson model individual effect is used >> as factor, but it is considered as nuisance term where parameter >> estimates are not needed. >> >> Could anyone point how absorbing individual effect could be carried >> out in glm. >> >> There is already code for Stata >> (http://statistics.open.ac.uk/sccs/stata/aglm.ado), Genstat >> (http://statistics.open.ac.uk/sccs/genstat%5Csccs.gen), Glim >> (http://statistics.open.ac.uk/sccs/glim/SCCS.MAC) , and SAS >> (http://statistics.open.ac.uk/sccs/sas/macro/poisreg.sas). >> >> >> Regards, >> Jari Haukka >> >> ______________________________________________ >> R-help at stat.math.ethz.ch mailing list >> https://stat.ethz.ch/mailman/listinfo/r-help >> PLEASE do read the posting guide >> http://www.R-project.org/posting-guide.html >> and provide commented, minimal, self-contained, reproducible code. >> >> >