Is the problem clear? If not please look at the following example based on
the real problem I'm dealing with:
Be TOT the total daily mortality, OBS a vector of sequential number
indicating the period (days) and MSO2 the daily value of a pollutant. I
want estimate the relation between TOT, on a side, and OBS and the
pollutant at various lags, on the other.
Moreover I want to reduce the subset used, imposing a max value for the
pollutant to exclude anomalous values.
As argument for the function
gaminq(model,pol,nlag,sbset)
I will give:
1) the core model formula (without pollutant):
model=TOT~lo(OBS,span=0.5)
2) the name of the pollutant: pol=MSO2
3) the max lag for the pollutant (if I give 2, I want to estimate
the model for lag 0,1 and 2): nlag=2
4) the subset: sbset=matinq[1:length(pol),(i+1)]<200
Where 'matinq' is the pollutant matrix, formed by the lagged values from 0
to nlag, so the first column is lag 0, the second is lag 1, and so on. And
'i' is the counter of a cicle like
for (i in 0:nlag)
{....}
Obviously I get an error message, because the program doesn't know yet the
variable 'matinq'.
I have thought to give the subset as a string and coerce it as a logical
expression after creating 'matinq', but I don't know how to do this.
What is your advice?
Thank you very much.
Gianluca Marchi
-----------------------------------------------------------------------
This message was distributed by s-news@wubios.wustl.edu. To unsubscribe
send e-mail to s-news-request@wubios.wustl.edu with the BODY of the
message: unsubscribe s-news