Jimmy Martina
2010-Nov-22 14:27 UTC
[R] What if geoRglm results showed that a non-spacial model fits?
Hi R-people: Working in geoRglm, it shows me, according to AIC criterion, that the non-spacial model describes the process in a better way. It's the first time that I'm facing up to. These are my results: OP2003Seppos.AICnonsp-OP2003Seppos.AICsp #[1] -4 (OP2003Seppos.lf0.p<-exp(OP2003Seppos.lf0$beta)/(1+exp(OP2003Seppos.lf0$beta))) #P non spatial #[1] 0.9717596 (OP2003Seppos.lf.p<-exp(OP2003Seppos.lf$beta)/(1+exp(OP2003Seppos.lf$beta))) #P spatial #[1] 0.9717596 It must what have an important influence at kriging, because it shows as following: OP2003Sepposbin.krig<-glsm.krige(OP2003Seppos.tune,loc=OP2003Seppospro.pred.grid,bor=OP2003Sepposbor) #glsm.krige: Prediction for a generalised linear spatial model #There are 50 or mode advices (use warnings() to see them) #> warnings() #Warning messages: #1: In asympvar(kpl.result$predict, messages = FALSE) ... : # value of argument lag.max is not suffiently long #2: In asympvar(kpl.result$predict, messages = FALSE) ... : # value of argument lag.max is not suffiently long Help me, please. [[alternative HTML version deleted]]
Jimmy Martina
2010-Nov-25 14:12 UTC
[R] What to do if geoRglm results showed that a non-spacial model fits better?
Hi R-floks: Working in geoRglm, it shows me, according to AIC criterion, that the non-spacial model describes the process in a better way. It's the first time that I'm facing up to. These are my results: OP2003Seppos.AICnonsp-OP2003Seppos.AICsp #[1] -4 (OP2003Seppos.lf0.p<-exp(OP2003Seppos.lf0$beta)/(1+exp(OP2003Seppos.lf0$beta))) #P non spatial #[1] 0.9717596 (OP2003Seppos.lf.p<-exp(OP2003Seppos.lf$beta)/(1+exp(OP2003Seppos.lf$beta))) #P spatial #[1] 0.9717596 It must what have an important influence at kriging, because it shows as following: OP2003Sepposbin.krig<-glsm.krige(OP2003Seppos.tune,loc=OP2003Seppospro.pred.grid,bor=OP2003Sepposbor) #glsm.krige: Prediction for a generalised linear spatial model #There are 50 or mode advices (use warnings() to see them) #> warnings() #Warning messages: #1: In asympvar(kpl.result$predict, messages = FALSE) ... : # value of argument lag.max is not suffiently long #2: In asympvar(kpl.result$predict, messages = FALSE) ... : # value of argument lag.max is not suffiently long I don't understand the warnings. Help me, please. [[alternative HTML version deleted]]