[R] completing missing samples

[Rich Shepard]

On Sat, 17 Sep 2011, R. Michael Weylandt wrote:

> How are your time samples missing?

Michael/Eran,

   Pardon me for jumping in since I did not read the original message, but I
have a more fundamental question: are the missing time samples meaningful?

   For example, a time series of precipitation should have strings of no
precipitation (even here in the Pacific Northwest), and the lack of
measurable precipitation is meaningful so those values should not be filled
in.

   Another example: you're supposed to collect water from a stream once per
week, but there are times when you cannot get there because the water level
is too high, there's too much snow on the ground to get to the stream, or
the stream has dried out in the summer. There are no valid ways to fill in
the missing samples that don't distort the results.

   So, just for my curiosity, why do you want to fill in those time slots?

Rich