mxFactor {OpenMx}R Documentation

Fail-safe Factors

Description

This is a wrapper for the R function factor.

OpenMx requires ordinal data to be ordered. R's factor function doesn't enforce this, hence this wrapper exists to throw an error should you accidentally try and run with ordered = FALSE.

Also, the ‘levels’ parameter is optional in R's factor function. However, relying on the data to specify the data is foolhardy for the following reasons: The factor function will skip levels missing from the data: Specifying these in levels leaves the list of levels complete. Data will often not explore the min and max level that the user knows are possible. For these reasons this function forces you to write out all possible levels explicitly.

Usage

mxFactor(x = character(), levels, labels = levels, 
    exclude = NA, ordered = TRUE, collapse = FALSE)

Arguments

x

either a vector of data or a data.frame object.

levels

a mandatory vector of the values that 'x' might have taken.

labels

_either_ an optional vector of labels for the levels, _or_ a character string of length 1.

exclude

a vector of values to be excluded from the set of levels.

ordered

logical flag to determine if the levels should be regarded as ordered (in the order given). Required to be TRUE.

collapse

logical flag to determine if duplicate labels should collapsed into a single level

Details

If ‘x’ is a data.frame, then all of the columns of ‘x’ are converted into ordered factors. If ‘x’ is a data.frame, then ‘levels’ and ‘labels’ may be either a list or a vector. When ‘levels’ is a list, then different levels are assigned to different columns of the constructed data.frame object. When ‘levels’ is a vector, then the same levels are assigned to all the columns of the data.frame object. The function will throw an error if ‘ordered’ is not TRUE or if ‘levels’ is missing. See factor for more information on creating ordered factors.

References

The OpenMx User's guide can be found at http://openmx.psyc.virginia.edu/documentation.

Examples

myVar <- c("s", "t", "a", "t", "i", "s", "t", "i", "c", "s")
ff    <- mxFactor(myVar, levels=letters)
# Note: letters is a built in list of all lowercase letters of the alphabet
ff
# [1] s t a t i s t i c s
# Levels: a < b < c < d < e < f < g < h < i < j < k < l < m < n < o < p < q <
#  r < s < t < u < v < w < x < y < z

as.integer(ff)  # the internal codes

factor(ff)      # NOTE: drops the levels that do not occur.
                # mxFactor prevents you doing this unintentionally.

# This example works on a dataframe
foo <- data.frame(x=c(1:3),y=c(4:6),z=c(7:9))

# Applys one set of levels to all three columns
mxFactor(foo, c(1:9))

# Apply unique sets of levels to each variable
mxFactor(foo, list(c(1:3), c(4:6), c(7:9)))


mxFactor(foo, c(1:9), labels=c(1,1,1,2,2,2,3,3,3), collapse=TRUE)


[Package OpenMx version 2.3.1 Index]