Impute loess
Imputes missing data or smooths using Loess regression
impute.loess(y, s = 0.2, smooth = FALSE)
y |
A vector to impute |
s |
Smoothing parameter () |
smooth |
(FALSE/TRUE) Smooth data, else only replace NA's |
Performs a local polynomial regression to smooth data or to impute NA values. The minimal number of non-NA observations to reliably impute/smooth values is 6. There is not a reliably way to impute NA's on the tails of the distributions so if the missing data is in the first or last position of the vector it will remain NA. Please note that smooth needs to be TRUE to return a smoothed vector, else only NA's will be imputed.
a vector the same length as x with NA values filled or the data smoothed (or both)..
Jeffrey S. Evans <jeffrey_evans<at>tnc.org>
data(cor.data) d <- cor.data[[1]][,2] plot(d, type="l") lines(impute.loess(d, s=0.3, smooth=TRUE), lwd=2, col="red") # add some NA's d <- d[1:100] d[sample(30:70, 5)] <- NA d impute.loess(d, s=0.2)
Please choose more modern alternatives, such as Google Chrome or Mozilla Firefox.