[R] please help for mgcv package

[Joshua Wiley]

On Tue, Jun 21, 2011 at 9:20 AM, Sarah Goslee <sarah.goslee at gmail.com> wrote:
> Hi,
>
> On Tue, Jun 21, 2011 at 11:27 AM, pigpigmeow <glorykwok at hotmail.com> wrote:
>> i read a book from WOOD, there's an example which is talking about the
>> pollutant.

Read the text carefully!!!

>>
>> library(gamair)
>> library(mgcv)
>
> data(chicago) # otherwise this isn't reproducible
>
>> y<-gam(death~s(time,bs="cr",k=200)+s(pm10median,bs="cr")+s(so2median,bs="cr")+s(o3median,bs="cr")+s(tmpd,bs="cr"),data=chicago,family=Possion)
>
> # this line won't work
> y<-gam(death~s(time,bs="cr",k=200)+s(pm10median,bs="cr")+s(so2median,bs="cr")+s(o3median,bs="cr")+s(tmpd,bs="cr"),data=chicago,family=poisson())
>
>
>> lag.sum<-function(a,10,11)

but this line is not what is in the book...I know it is not the
easiest to see, but those are ls not 1s.

>> {n<-length(a)
>> b<-rep(0,n-11)
>> for(i in 0:(11-10))
>> b<-b+a[(i+1):(n-11+i)]

At best this risks being confusing, at worst it could lead to errors.
If you use a for loop, it is good practice not to break the line (what
the good Dr. Wood does), or if you prefer to break the line, wrap it
in curly braces {}.  See ?for for the documentation.
>>
>> b}

Here is a copy (following my stylistic preferences, but should be
identical in function):

lag.sum <- function(a, l0, l1) {
  n <- length(a)
  b <- rep(0, n - l1)
  for (i in 0:(l1 - l0)) {
    b <- b + a[(i + 1):(n - l1 + i)]
  }
  b
}


Regarding what rep() does, see the documentation, ?rep  also, play
with simple examples to see its behavior:

rep(1, 10)
rep(1:3, 5)

If you are unfamiliar with R, you might have an easier go of it if you
read some of the myriad of online documentation/tutorials (or a book)
designed to introduce you to R.  GAMs is not really written to
introduce R, so it will be unnecessarily difficult to learn R from it.
 Here is one of the official manuals:

http://cran.r-project.org/doc/manuals/R-intro.html

Patrick Burns also has a number of tutorials for different levels of useRs:

http://www.burns-stat.com/pages/tutorials.html

HTH,

Josh

>
> Why would you pass an unnamed constant to a function, and then try to
> change them later? When you use lag.sum(), you try to set those
> arguments to other values.
>
> lag.sum<-function(a1, a2, a3)
> {
> n<-length(a1)
> b<-rep(0, n-a3)
> for(i in 0:(a2-a3))
> b<- b + a1[(i+1):(n-a3+i)]
>
> b
> }
>
>
>> death<-chicago$death[4:5114]
>> time<-chicago$time[4:5114]
>> o3<-lag.sum(chicago$o3median,0,3)
>> tmp<-lag.sum(chicago$tmpd,0,3)
>
> Like here. Your function hard-coded the values in, or tried to, and
> yet you want 0 and 3 now instead of 10 and 11.
>
>> pm10<-lag.sum(log(chicago$pm10median+40),0,3)
>> so2<-lag.sum(log(chicago$so2median+10),0,)
>>
>> I don't know what is the script (Bold font ) used for......
>
> This is a plain-text list. Bold font doesn't show up.
>
>> and it shows "Error: unexpected numeric constant in "lag.sum<-function(a,10"
>> ", why?
>>
>> anyone can answer me?
>>
>
> I don't have the book you reference, but I wonder if that's really what it says.
>
> Sarah
>
>
> --
> Sarah Goslee
> http://www.functionaldiversity.org
>
> ______________________________________________
> R-help at r-project.org 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.
>



-- 
Joshua Wiley
Ph.D. Student, Health Psychology
University of California, Los Angeles
http://www.joshuawiley.com/