reorders: Reorder Levels of Factor Data
In quest: Prepare Questionnaire Data for Analysis

Description Usage Arguments Value See Also Examples

reorders re-orders the levels of factor data. The factors are columns in a data.frame where the same reordering scheme is desired. This is often useful before using factor data in a statistical analysis (e.g., lm) or a graph (e.g., ggplot). It is essentially a vectorized version of reorder.default.

1	reorders(data, fct.nm, ord.nm = NULL, fun, ..., suffix = "_r")

`data`	data.frame of data.
`fct.nm`	character vector of colnames in `data` that specify the factor columns. If any of the columns specified by `fct.nm` are not factors, then an error is returned.
`ord.nm`	character vector of length 1 or `NULL`. If a character vector of length 1, it is a colname in `data` specifying the column in `data` that will be used in conjunction with `fun` to re-order the factor columns. If `NULL` (default), it is assumed that each factor column itself will be used in conjunction with `fun` to re-order the factor columns.
`fun`	function that will be used to re-order the factor columns. The function is expected to input an atomic vector of length = `nrow(data)` and return an atomic vector of length 1. `fun` is applied to `data[[ord.nm]]` if `ord.nm` is a character vector of length 1 or applied to each column in `data[fct.nm]` if `ord.nm` = `NULL`.
`...`	additional named arguments used by `fun`. For example, if `fun` is `mean`, the user might specify an argument `na.rm = TRUE` to set the `na.rm` argument in the `mean` function.
`suffix`	character vector of length 1 specifying the string that will be appended to the end of the colnames in the return object.

data.frame of re-ordered factor columns with colnames = paste0(fct.nm, suffix).

reorder.default

# factor vector
reorder(x = state.region, X = state.region,
   FUN = length) # least frequent to most frequent
reorder(x = state.region, X = state.region,
   FUN = function(vec) {-1 * length(vec)}) # most frequent to least frequent

# data.frame of factors
infert_fct <- infert
fct_nm <- c("education","parity","induced","case","spontaneous")
infert_fct[fct_nm] <- lapply(X = infert[fct_nm], FUN = as.factor)
x <- reorders(data = infert_fct, fct.nm = fct_nm,
   fun = length) # least frequent to most frequent
lapply(X = x, FUN = levels)
y <- reorders(data = infert_fct, fct.nm = fct_nm,
   fun = function(vec) {-1 * length(vec)}) # most frequent to least frequent
lapply(X = y, FUN = levels)
# ord.nm specified as a different column in data.frame
z <- reorders(data = infert_fct, fct.nm = fct_nm, ord.nm = "pooled.stratum",
   fun = mean) # category with highest mean for pooled.stratum to
   # category with lowest mean for pooled.stratum
lapply(X = z, FUN = levels)