first: Identify Elements Within Cells Meeting Some Instance of a...

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

Description

Within each level of within, find the n^th element for which where is TRUE, and repeat it at all positions in the level.

Usage

nth  (x,where,within,n=1,...)
first(x,where,within,...)
last (x,where,within,...)
only (x,where,within,...)
distance(where,within,n=1,...)
before(  where,within,n=1,...)
at(      where,within,n=1,...)
after(   where,within,n=1,...)

Arguments

x	vector (possibly factor)
where	vector, coerced to logical
within	a vector or list of vectors serving as an index
n	an integer giving the instance of interest
...	ignored arguments.

Details

nth returns, for each position in x, the n^th element of x, optionally limiting candidate elements by where, and optionally breaking the evaluation across subsets as specified by within. n can be 0, returning all NA, or negative, which counts instances from the end of the vector or subsets. If n is NA, all elements are returned. NA is returned at positions for which the condition where is not met n times.

All of x, where, and within are optional. The ‘scale’ of the function is set to the longest of the three (member-wise for within if it is a list). Missing x is replaced with the subscripts implied by ‘scale’. Missing where and within are replaced with TRUE, repeated to length 'scale'. No warnings are given if within has NA, or list members with different lengths.

first and last are convenience wrappers for nth that set n to 1 and -1, respectively. only is like first, but returns NA if a second instance exists.

distance returns subscripts less the nth subscripts, optionally considering where and within. before, at, and after test whether distance is less than, equal to, or greater than zero.

Value

vector of same class as x, if supplied, else of class integer. Logical for before, at, and after.

Author(s)

Tim Bergsma

References

http://metrumrg.googlecode.com

See Also

• match

Examples

first(1:10)
first(1:10,within=rep(1:2,each=5))
last( 1:10,within=rep(1:2,each=5))
first(1:10,where=0:1,within=rep(1:2,each=5))
test <- data.frame(
	id=c(1,1,1,1,1,2,2,2,2,2),
	time=c(1,2,3,4,5,1,2,3,4,5),
	evid=c(0,1,0,1,0,1,0,1,0,1)
)
test$secondDose  <- with(test,nth(where=evid==1, n=2,within=id))
test$lastDose    <- with(test,nth(where=evid==1,n=-1,within=id))
test$beforeDose2 <- with(test,before(where=evid==1, n=2,within=id))
test$tafd        <- with(test,time - first(time,where=evid==1,within=id))#time after first dose
test
#    id time evid secondDose lastDose beforeDose2 tafd
# 1   1    1    0          4        4        TRUE   -1
# 2   1    2    1          4        4        TRUE    0
# 3   1    3    0          4        4        TRUE    1
# 4   1    4    1          4        4       FALSE    2
# 5   1    5    0          4        4       FALSE    3
# 6   2    1    1          8       10        TRUE    0
# 7   2    2    0          8       10        TRUE    1
# 8   2    3    1          8       10       FALSE    2
# 9   2    4    0          8       10       FALSE    3
# 10  2    5    1          8       10       FALSE    4
only(
	1:9,
	where= c(0,1,0,  1,0,1,  0,0,0),
	within=c(1,1,1,  2,2,2,  3,3,3)
)
#[1]  2  2  2 NA NA NA NA NA NA

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