pasteByRowOrdered: Paste data.frame rows into an ordered factor

In jmw86069/jamba: Just Analysis Methods Base

Paste data.frame rows into an ordered factor

Description

Paste data.frame rows into an ordered factor

Usage

pasteByRowOrdered(
  x,
  sep = "_",
  na.rm = TRUE,
  condenseBlanks = TRUE,
  includeNames = FALSE,
  keepOrder = FALSE,
  byCols = seq_len(ncol(x)),
  na.last = TRUE,
  ...
)

Arguments

x	data.frame
sep	character separator to use between columns
na.rm	logical whether to remove NA values, or include them as "NA"
condenseBlanks	logical whether to condense blank or empty values without including an extra delimiter between columns.
includeNames	logical whether to include the colname delimited prior to the value, using sepName as the delimiter.
keepOrder	logical indicating whether non-factor columns should order factor levels based upon the existing order of unique items. This option is intended for data.frame whose columns are already sorted in proper order, but where columns are not factor with appropriate factor levels. Note that even when keepOrder=TRUE all existing factor columns will honor the order of factor levels already present in those columns.
byCols	integer or character passed to mixedSortDF(). This argument defines the order of columns sorted by mixedSortDF(), and does not affect the order of columns pasted. Columns are always pasted in the same order they appear in x. This argument byCols was previously passed via ... but is added here to make this connection more direct.
na.last	logical passed to base::factor() to determine whether NA values are first or last in factor level order.
...	additional arguments are passed to jamba::pasteByRow(), and to jamba::mixedSortDF().

Details

This function is an extension to jamba::pasteByRow() which pastes rows from a data.frame into a character vector. This function defines factor levels by running jamba::mixedSortDF(unique(x)) and calling jamba::pasteByRow() on the result. Therefore the original order of the input x is maintained while the factor levels are based upon the appropriate column-based sort.

Note that the ... additional arguments are passed to jamba::mixedSortDF() to customize the column-based sort order, used to define factor levels. A good way to test the order of factors is to run jamba::mixedSortDF(unique(x)) with appropriate arguments, and confirm the rows are ordered as expected.

Note also that jamba::mixedSortDF() uses jamba::mixedSort() which itself performs alphanumeric sort in order to keep values in proper numeric order where possible.

Value

factor vector whose levels are defined by existing factor levels, then by sorted values.

See Also

Other jam string functions: asSize(), breaksByVector(), fillBlanks(), formatInt(), gsubOrdered(), gsubs(), makeNames(), nameVector(), nameVectorN(), padInteger(), padString(), pasteByRow(), sizeAsNum(), tcount(), ucfirst()

Examples

f <- LETTERS;
df <- data.frame(A=f[rep(1:3, each=2)],
   B=c(NA, f[3]),
   C=c(NA, NA, f[2]))
df

# note that output is consistent with mixedSortDF()
jamba::mixedSortDF(df)
jamba::pasteByRowOrdered(df)

jamba::mixedSortDF(df, na.last=FALSE)
jamba::pasteByRowOrdered(df, na.last=FALSE)

jamba::mixedSortDF(df, byCols=c(3, 2, 1))
jamba::pasteByRowOrdered(df, byCols=c(3, 2, 1))

df1 <- data.frame(group=rep(c("Control", "ABC1"), each=6),
   time=rep(c("Hour2", "Hour10"), each=3),
   rep=paste0("Rep", 1:3))
# default will sort each column alphanumerically
pasteByRowOrdered(df1)

# keepOrder=TRUE will honor existing order of character columns
pasteByRowOrdered(df1, keepOrder=TRUE)

jmw86069/jamba documentation built on Feb. 28, 2025, 9:07 p.m.