[R] Re: bootstrap and nls

[Spencer Graves]

      I would think that having to eliminate close to 10% of bootstrap 
samples could be an issue. 

      Have you considered using
   

	  myboot<-boot(Raies,dif.param,R=1000, sim="permutation")


      Permutation testing does sampling without replacement rather than 
sampling with replacement.  Perhaps someone else can enlighten us both 
on the arguments for and against both permutation testing and 
bootstrapping, but permutation testing could eliminate one source of NAs 
in the output;  in any event, I would want to keep track of the percent 
of NAs obtained under both.  For a statistical audience, you would need 
to cite literature explaining the difference between permutation testing 
and bootsrapping, to help people understand your results. 

      hope this helps.  spencer graves

Patrick Giraudoux wrote:

>Hi,
>
>Just a recap on the trials done based on Spencer Grave, Bervin A Turlach and Christian Ritz's advise.
>
>On their suggestion, the trouble mentioned has well been turned using the function try() (using the algorithm "plinear" unstead of
>"default" was unsuccessful) in the following way:
>
>library(boot)
>
>dif.param<-function(data,i){
>RaiesLossA.nls<-try(nls(SolA~a/(1+b*Tps)^c,start=c(a=31,b=0.5,c=0.6),data[i,]),silent=TRUE)
>RaiesLossB.nls<-try(nls(Solb~a/(1+b*Tps)^c,start=c(a=33,b=1.4,c=0.5),data[i,]),silent=TRUE)
>
>if ( (inherits(RaiesLossA.nls,"try-error")) |
>(inherits(RaiesLossB.nls,"try-error"))) {return(NA)} else {
>return(RaiesLossA.nls$m$getPars()-RaiesLossB.nls$m$getPars())}
>
>}
>
>myboot<-boot(Raies,dif.param,R=1000)
>
>
>The boot objet "myboot" that one get cannot however be handled, eg with boot.ci(),  because "myboot$t" has NA values. This can be
>corrected in this way:
>
>myboot$t<-na.omit(myboot$t)
>myboot$R<-length(myboot$t[,1])
>
>boot.ci(myboot,index=1, type=c("norm","basic","perc", "bca"))
>
>Studentized CI appear to be quite unstable here and were not used.
>
>That's it!
>
>The effect of omitting some bootstrap replicates (the less than 10% that could not lead to a propper fit)  is however questionable:
>may this
>biase the result in a way? I am not too much worried about it for my current purpose (crude comparison of parameters), but I
>guess that it may be an issue for true statisticians...
>
>Many thanks for the most helpful hints provided,
>
>Patrick Giraudoux
>
>
>
>
>  
>
>>Hi,
>>
>>I am trying to bootstrap the difference between each parameters among two non linear regression (distributed loss model) as
>>following:
>>
>># data.frame
>>
>>
>>    
>>
>>>Raies[1:10,]
>>>
>>>
>>>      
>>>
>>  Tps  SolA  Solb
>>1    0 32.97 35.92
>>2    0 32.01 31.35
>>3    1 21.73 22.03
>>4    1 23.73 18.53
>>5    2 19.68 18.28
>>6    2 18.56 16.79
>>7    3 18.79 15.61
>>8    3 17.60 13.43
>>9    4 14.83 12.76
>>10   4 17.33 14.91
>>etc...
>>
>># non linear model (work well)
>>
>>RaiesLossA.nls<-nls(SolA~a/(1+b*Tps)^c,start=c(a=32,b=0.5,c=0.6))
>>RaiesLossB.nls<-nls(Solb~a/(1+b*Tps)^c,start=c(a=33,b=1.5,c=0.5))
>>
>>
>># bootstrap
>>library(boot)
>>
>>dif.param<-function(data,i){
>>RaiesLossA.nls<-nls(SolA[i]~a/(1+b*Tps[i])^c,start=c(a=31,b=0.5,c=0.6))
>>RaiesLossB.nls<-nls(Solb[i]~a/(1+b*Tps[i])^c,start=c(a=33,b=1.4,c=0.5))
>>RaiesLossA.nls$m$getPars()-RaiesLossB.nls$m$getPars()
>>}
>>
>>
>>
>>    
>>
>>>myboot<-boot(Raies,dif.param,R=1000)
>>>
>>>
>>>      
>>>
>>Error in numericDeriv(form[[3]], names(ind), env) :
>>       Missing value or an Infinity produced when evaluating the model
>>
>>It seems that the init values (start=) may come not to be suitable while bootstraping. Data can be sent offline to whom wanted to
>>try on the dataset.
>>
>>Any hint welcome!
>>
>>Best regards,
>>
>>Patrick Giraudoux
>>    
>>
>
>
>
>
>
>
>  
>