funData (version 1.3-9)

extractObs: Extract observations of functional data

Description

This function extracts one or more observations and/or observations on a part of the domain from a funData, irregFunData or multiFunData object.

Usage

extractObs(
  object,
  obs = seq_len(nObs(object)),
  argvals = funData::argvals(object)
)
# S4 method for funData
subset(x, obs = seq_len(nObs(x)), argvals = funData::argvals(x))
# S4 method for multiFunData
subset(x, obs = seq_len(nObs(x)), argvals = funData::argvals(x))
# S4 method for irregFunData
subset(x, obs = seq_len(nObs(x)), argvals = funData::argvals(x))
# S4 method for funData,ANY,missing,missing
[(x, i, j, ..., drop = TRUE)
# S4 method for multiFunData,ANY,missing,missing
[(x, i, j, ..., drop = TRUE)
# S4 method for irregFunData,ANY,missing,missing
[(x, i = seq_len(nObs(x)), j, ..., drop = TRUE)

Value

An object of class funData, irregFunData or

multiFunData containing the desired observations.

Arguments

object

An object of class funData, irregFunData or multiFunData.

obs

A numeric vector, giving the indices of the observations to extract (default: all observations).

argvals

The part of the domain to be extracted (default: the whole domain object@argvals). Must be a list or a numeric vector (only for one-dimensional domains, see also the definition of funData, multiFunData).

x

An object of class funData, irregFunData or multiFunData (for subset).

i

A numeric vector, giving the indices of the observations to extract when using x[i]. Defaults to all observations.

j, drop

not used

...

Used to pass further arguments to extractObs. Here only usable for argvals.

Functions

• x[i:

Warning

The function is currently implemented only for functional data with up to three-dimensional domains.

Alias

The function subset is an alias for extractObs.

Details

In case of an irregFunData object, some functions may not have observation points in the given part of the domain. In this case, the functions are removed from the extracted dataset and a warning is thrown.

If only observations are to be extracted, the usual notation object[1:3] is equivalent to extractObs(object, obs = 1:3). This works only if the domain remains unchanged.

See Also

funData, irregFunData, multiFunData

Examples

# Univariate - one-dimensional domain
object1 <- funData(argvals = 1:5, X = rbind(1:5, 6:10))
extractObs(object1, obs = 1)
extractObs(object1, argvals = 1:3)
extractObs(object1, argvals = list(1:3)) # the same as the statement before
# alias
subset(object1, argvals = 1:3)

# Univariate - two-dimensional domains
object2 <- funData(argvals = list(1:5, 1:6), X = array(1:60, dim = c(2, 5, 6)))
extractObs(object2, obs = 1)
extractObs(object2, argvals = list(1:3, c(2,4,6))) # argvals must be supplied as list

# Univariate - irregular
irregObject <- irregFunData(argvals = list(1:5, 2:4), X = list(2:6, 3:5))
extractObs(irregObject, obs = 2)
extractObs(irregObject, argvals = 1:3)
extractObs(irregObject, argvals = c(1,5)) # throws a warning, as second function has no observations

# Multivariate
multiObject <- multiFunData(object1, object2)
extractObs(multiObject, obs = 2)
multiObject[2] # shorthand
extractObs(multiObject, argvals = list(1:3, list(1:3, c(2,4,6))))


### Shorthand via "[]"
object1[1]
object1[argvals = 1:3]
object2[1] 
object2[argvals = list(1:3, c(2,4,6))]
irregObject[2]
irregObject[argvals = 1:3]