Hello,

I am new to linear prediction. What I want to do is exactly described in the paper by Rybicki and Press (Interpolation, realization, and reconstruction of noisy, irregularly sampled data”

In essence, I have missing data and I want to reconstruct the time series. Is there anyway to use the routines in NR C++ to accomplish what was done the the RP paper? I would think the memcof procedure but I would also figure that there would be some method to indicate wich time periods were missing from the data set.

In terms of the power spectrum of unevenly sampled data….anyway to convert power spectrum back to time series?