inner.data.frame: Limit Data to Inner Quantiles by Imputing NA.

In metrumrg: Pharmacometric Tools for Data Preparation, Modeling, Simulation, and Reporting

Description

inner is generic. inner.data.frame imputes NA for cells within columns that represent extreme quantiles. By default, the ‘inner’ 95 percent of each column is preserved.

Usage

## S3 method for class 'data.frame'
inner(
	x,
	prob=0.95,
	tail=0.5*(1-prob),
	lo=tail,
	hi=prob+tail,
	include.lowest=TRUE,
	include.highest=TRUE,
	preserve=character(0),
	id.var=character(0),
	measure.var=setdiff(names(x),c(preserve,id.var)),
	na.rm=FALSE,
	...
)

Arguments

x	data.frame
prob	the fraction of data to preserve
tail	the fraction of data to ignore at each extreme
lo	the probability below which data will be ignored
hi	the probability above which data will be ignored
include.lowest	whether to preserve values at probability equal to lo
include.highest	whether to preserve values at probability equal to hi
preserve	vector of names for columns to preserve but ignore
id.var	vector of names for columns that indicate data subsets
measure.var	vector of names for columns to limit
na.rm	passed to quantile
...	passed to fun.aggregate

Details

prob and tail are not actually used internally, but serve only to calculate symmetric defaults for lo and hi. If the latter are supplied, the former are ignored. Tails need not be symmetric.

By default, all columns are classified as measure.var: an attempt will be made to limit such. Columns classified as preserve will simply be passed through to the result. Columns classified as id.var specify row subsets that are limited independently of eachother. See examples.

Value

a data frame with the same rows, columns, row order, and column order as x (except for dropped columns)

Author(s)

Tim Bergsma

References

http://metrumrg.googlecode.com

See Also

• helpPrepare

Examples

x <- airquality[c(1:10,32:41),]
x
inner(x,id.var=c('Month','Day'),na.rm=TRUE)#identity
inner(x,id.var='Month',preserve='Day',na.rm=TRUE)#quantiles within Month
inner(x,preserve=c('Month','Day'),na.rm=TRUE)#quantiles across all rows
inner(x,measure.var=c('Ozone','Solar.R','Wind','Temp'),na.rm=TRUE)# dropping Month, Day
inner(rock,prob=0.5)

metrumrg documentation built on May 2, 2019, 5:55 p.m.