Generates a lag one auto-regressive time series, where the first and last
values are fixed. Marginal expected value is zero and variance is one.
Generated values have a normal conditional distribution. This version replaces
genmissing and is much faster for long sample gaps.
Usage
genmissing(X1, XN, rho, N, Lfull)
Arguments
- X1
value before the gap
- XN
value after the gap
- rho
the lag one autocorrelation
- N
the length of the sequence including X1 and XN. It
is two more than the gap length
- Lfull
Cholesky factor created with precompute_L
Value
numeric vector of length N, conditioned on the
first value (X1) and last value (XN) with the specified lag one autocorrelation
in the limit (where N is large) the values are normal with mean 0 and variance 1
