I am using the function boot (library boot) to calculate bootstrap distributions
of parameters describing population dynamics. In rare occasions the resampling
can produce invalid values (caused by a division by zero). These values have no
biological importance and it seems like I have to exclude them in order to
calculate confidence intervals (function boot.ci). My questions are: (1) Am I
correct that the functions boot and boot.ci do not account for NA-values? (2) Is
it legitimate to substitute NA-values in boot$t with separately calculated
values and afterwards use boot.ci to calculate confidence intervals, or are
there other parts of the boot object used that would need adjusting?
Thanks in advanced for any comments
Sebastian Meyer
_________________________________
Sebastian T. Meyer (PhD-student)
Department for Plant Ecology and Systematics
University of Kaiserslautern
PO-Box 3049 | 67653 Kaiserslautern | Germany
Fone: ++49 631/205-4402
Fax: ++49 631 205 2998
meyerst at rhrk.uni-kl.de
