Stateful recursive-least-squares adaptive filter
Usage
rlsFilter(n_taps, forgetting = 0.99, delta = 1, initial_weights = NULL)
Arguments
- n_taps
Exact adaptive-filter order.
- forgetting
Forgetting factor in (0, 1].
- delta
Positive initial inverse-covariance scale.
- initial_weights
Optional finite n_taps by channel matrix.
Value
A mutable RLSFilter streaming processor.
Examples
set.seed(1)
reference <- cos(2 * pi * 6 * seq_len(150) / 100)
signal <- 0.5 * reference + rnorm(150, sd = 0.05)
filt <- rlsFilter(n_taps = 3L, forgetting = 0.995)
result <- update(filt, matrix(signal, ncol = 1),
reference = matrix(reference, ncol = 1))
length(result$output)
#> [1] 150