data.table (version 1.15.4) Fast melt for data.table


melt is data.table's wide-to-long reshaping tool. We provide an S3 method for melting data.tables. It is written in C for speed and memory efficiency. Since v1.9.6, allows melting into multiple columns simultaneously.


## fast melt a data.table
# S3 method for data.table
melt(data, id.vars, measure.vars, = "variable", = "value",
    ..., na.rm = FALSE, variable.factor = TRUE,
    value.factor = FALSE,
    verbose = getOption("datatable.verbose"))


An unkeyed data.table containing the molten data.



A data.table object to melt.


vector of id variables. Can be integer (corresponding id column numbers) or character (id column names) vector. If missing, all non-measure columns will be assigned to it. If integer, must be positive; see Details.


Measure variables for melting. Can be missing, vector, list, or pattern-based.

  • When missing, measure.vars will become all columns outside id.vars.

  • Vector can be integer (implying column numbers) or character (column names).

  • list is a generalization of the vector version -- each element of the list (which should be integer or character as above) will become a melted column.

  • Pattern-based column matching can be achieved with the regular expression-based patterns syntax; multiple patterns will produce multiple columns.

For convenience/clarity in the case of multiple melted columns, resulting column names can be supplied as names to the elements measure.vars (in the list and patterns usages). See also Examples.

name (default 'variable') of output column containing information about which input column(s) were melted. If measure.vars is an integer/character vector, then each entry of this column contains the name of a melted column from data. If measure.vars is a list of integer/character vectors, then each entry of this column contains an integer indicating an index/position in each of those vectors. If measure.vars has attribute variable_table then it must be a data table with nrow = length of measure.vars vector(s), each row describing the corresponding measured variables(s), (typically created via measure) and its columns will be output instead of the column.

name for the molten data values column(s). The default name is 'value'. Multiple names can be provided here for the case when measure.vars is a list, though note well that the names provided in measure.vars take precedence.


If TRUE, NA values will be removed from the molten data.


If TRUE, the variable column will be converted to factor, else it will be a character column.


If TRUE, the value column will be converted to factor, else the molten value type is left unchanged.


TRUE turns on status and information messages to the console. Turn this on by default using options(datatable.verbose=TRUE). The quantity and types of verbosity may be expanded in future.


any other arguments to be passed to/from other methods.


If id.vars and measure.vars are both missing, all non-numeric/integer/logical columns are assigned as id variables and the rest as measure variables. If only one of id.vars or measure.vars is supplied, the rest of the columns will be assigned to the other. Both id.vars and measure.vars can have the same column more than once and the same column can be both as id and measure variables. also accepts list columns for both id and measure variables.

When all measure.vars are not of the same type, they'll be coerced according to the hierarchy list > character > numeric > integer > logical. For example, if any of the measure variables is a list, then entire value column will be coerced to a list.

From version 1.9.6, melt gains a feature with measure.vars accepting a list of character or integer vectors as well to melt into multiple columns in a single function call efficiently. If a vector in the list contains missing values, or is shorter than the max length of the list elements, then the output will include runs of missing values at the specified position, or at the end. The function patterns can be used to provide regular expression patterns. When used along with melt, if cols argument is not provided, the patterns will be matched against names(data), for convenience.

Attributes are preserved if all value columns are of the same type. By default, if any of the columns to be melted are of type factor, it'll be coerced to character type. To get a factor column, set value.factor = TRUE. also preserves ordered factors.

Historical note: was originally designed as an enhancement to reshape2::melt in terms of computing and memory efficiency. reshape2 has since been superseded in favour of tidyr, and melt has had a generic defined within data.table since v1.9.6 in 2015, at which point the dependency between the packages became more etymological than programmatic. We thank the reshape2 authors for the inspiration.

See Also


Run this code
DT <- data.table(
  i_1 = c(1:5, NA),
  n_1 = c(NA, 6, 7, 8, 9, 10),
  f_1 = factor(sample(c(letters[1:3], NA), 6L, TRUE)),
  f_2 = factor(c("z", "a", "x", "c", "x", "x"), ordered=TRUE),
  c_1 = sample(c(letters[1:3], NA), 6L, TRUE),
  c_2 = sample(c(LETTERS[1:2], NA), 6L, TRUE),
  d_1 = as.Date(c(1:3,NA,4:5), origin="2013-09-01"),
  d_2 = as.Date(6:1, origin="2012-01-01")
# add a couple of list cols
DT[, l_1 := DT[, list(c=list(rep(i_1, sample(5, 1L)))), by = i_1]$c]
DT[, l_2 := DT[, list(c=list(rep(c_1, sample(5, 1L)))), by = i_1]$c]

# id.vars, measure.vars as character/integer/numeric vectors
melt(DT, id.vars=1:2, measure.vars="f_1")
melt(DT, id.vars=c("i_1", "n_1"), measure.vars=3) # same as above
melt(DT, id.vars=1:2, measure.vars=3L, value.factor=TRUE) # same, but 'value' is factor
melt(DT, id.vars=1:2, measure.vars=3:4, value.factor=TRUE) # 'value' is *ordered* factor

# preserves attribute when types are identical, ex: Date
melt(DT, id.vars=3:4, measure.vars=c("d_1", "d_2"))
melt(DT, id.vars=3:4, measure.vars=c("n_1", "d_1")) # attribute not preserved

# on list
melt(DT, id.vars=1, measure.vars=c("l_1", "l_2")) # value is a list
  melt(DT, id.vars=1, measure.vars=c("c_1", "l_1")) # c1 coerced to list, with warning

# on character
melt(DT, id.vars=1, measure.vars=c("c_1", "f_1")) # value is char
  melt(DT, id.vars=1, measure.vars=c("c_1", "n_1")) # n_1 coerced to char, with warning

# on na.rm=TRUE. NAs are removed efficiently, from within C
melt(DT, id.vars=1, measure.vars=c("c_1", "c_2"), na.rm=TRUE) # remove NA

# measure.vars can be also a list
# melt "f_1,f_2" and "d_1,d_2" simultaneously, retain 'factor' attribute
# convenient way using internal function patterns()
melt(DT, id.vars=1:2, measure.vars=patterns("^f_", "^d_"), value.factor=TRUE)
# same as above, but provide list of columns directly by column names or indices
melt(DT, id.vars=1:2, measure.vars=list(3:4, c("d_1", "d_2")), value.factor=TRUE)
# same as above, but provide names directly:
melt(DT, id.vars=1:2, measure.vars=patterns(f="^f_", d="^d_"), value.factor=TRUE)

# na.rm=TRUE removes rows with NAs in any 'value' columns
melt(DT, id.vars=1:2, measure.vars=patterns("f_", "d_"), value.factor=TRUE, na.rm=TRUE)

# 'na.rm=TRUE' also works with list column, but note that only
# returns TRUE if the list element is a length=1 vector with an NA., two.NA=c(NA,NA)))
melt(DT, id.vars=1:2, measure.vars=patterns("l_", "d_"), na.rm=FALSE)
melt(DT, id.vars=1:2, measure.vars=patterns("l_", "d_"), na.rm=TRUE)

# measure list with missing/short entries results in output with runs of NA
DT.missing.cols <- DT[, .(d_1, d_2, c_1, f_2)]
melt(DT.missing.cols, measure.vars=list(d=1:2, c="c_1", f=c(NA, "f_2")))

# specifying columns to melt via separator.
melt(DT.missing.cols, measure.vars=measure(, number=as.integer, sep="_"))

# specifying columns to melt via regex.
melt(DT.missing.cols, measure.vars=measure(, number=as.integer, pattern="(.)_(.)"))

Run the code above in your browser using DataLab