\HeaderA{na}{Basic Statistical Functions for Handling Missing Values}{na}
\aliasA{cor.na}{na}{cor.na}
\aliasA{length.na}{na}{length.na}
\aliasA{log.na}{na}{log.na}
\aliasA{mean.na}{na}{mean.na}
\aliasA{order.na}{na}{order.na}
\aliasA{prod.na}{na}{prod.na}
\aliasA{quantile.na}{na}{quantile.na}
\aliasA{scale.na}{na}{scale.na}
\aliasA{sum.na}{na}{sum.na}
\aliasA{var.na}{na}{var.na}
\keyword{arith}{na}
\begin{Description}\relax
Basic statistical functions for handling missing values or NA. \\ 
In \code{log.na}, \code{sum.na}, \code{mean.na} and \code{var.na},
\code{quantile.na}, \code{length.na}, missing values are omitted
from the calculation. \\ 
The function \code{cor.na} calls \code{cor} with the argument
\code{use="pairwise.complete.obs"}. \\ 
The function \code{order.na} only handles vector arguments and not
lists.  However, it gives the option of omitting the NAs
(\code{na.last=NA}), of placing the NAs at the start of the ordered
vector (\code{na.last=F}) or at the end (\code{na.last=T}). \\ 
The function \code{scale.na} is a modified version of
\code{\LinkA{scale}{scale}} which allows NAs in the variance calculation. If
\code{scale = T}, the function \code{f} in \code{scale.na} uses
\code{var.na} to perform the variance calculation.
The function \code{prod.na} is similar to the \code{\LinkA{prod}{prod}}
function with \code{na.rm=TRUE}. This function returns the product of
all the values present in its arguments, omitting any missing values.
\end{Description}
\begin{Author}\relax
Yee Hwa Yang, \email{jean@biostat.berkeley.edu} \\
\end{Author}
\begin{SeeAlso}\relax
\code{\LinkA{log}{log}}, \code{\LinkA{sum}{sum}}, \code{\LinkA{mean}{mean}},
\code{\LinkA{var}{var}}, \code{\LinkA{cor}{cor}}, \code{\LinkA{order}{order}},
\code{\LinkA{scale}{scale}}, \code{\LinkA{prod}{prod}}.
\end{SeeAlso}