Hi all,
I have observations that mix together contributions from several smooth
functions of time. These smooths represent fluxes on a graph. Some of the
observations are centered on the edges of the graph, and directly measure one
flux. The others measurements are centered on the nodes of the graph and these
rows of the problem relate sources and sinks on the nodes to the sum of
surrounding fluxes. So they mix together multiple smooths.
Is there a straightforward way to handle combinations of smooths on the same
covariate in a tool such as mgcv? I have come up with some crude ideas like
defining time1, time2, time3 and then combine them with indicator variables:
y ~I(1)s(time1) + I(2)s(time2) + ....
Is this safe? Is there a better way to handle different smooths over the same
covariate? Some trick with factors?
Thanks much!
[[alternative HTML version deleted]]
______________________________________________
R-help@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.
