format
Encode in a Common Format
Format an R object for pretty printing.
Usage
format(x, …)
# S3 method for default
format(x, trim = FALSE, digits = NULL, nsmall = 0L,
justify = c("left", "right", "centre", "none"),
width = NULL, na.encode = TRUE, scientific = NA,
big.mark = "", big.interval = 3L,
small.mark = "", small.interval = 5L,
decimal.mark = getOption("OutDec"),
zero.print = NULL, drop0trailing = FALSE, …)
# S3 method for data.frame
format(x, …, justify = "none")
# S3 method for factor
format(x, …)
# S3 method for AsIs
format(x, width = 12, …)
Arguments
 x
 any R object (conceptually); typically numeric.
 trim
 logical; if
FALSE
, logical, numeric and complex values are rightjustified to a common width: ifTRUE
the leading blanks for justification are suppressed.  digits
 how many significant digits are to be used for
numeric and complex
x
. The default,NULL
, usesgetOption("digits")
. This is a suggestion: enough decimal places will be used so that the smallest (in magnitude) number has this many significant digits, and also to satisfynsmall
. (For the interpretation for complex numbers seesignif
.)  nsmall
 the minimum number of digits to the right of the decimal
point in formatting real/complex numbers in nonscientific formats.
Allowed values are
0 <= nsmall <= 20
.  justify
 should a character vector be leftjustified (the default), rightjustified, centred or left alone. Can be abbreviated.
 width
default
method: the minimum field width orNULL
or0
for no restriction.AsIs
method: the maximum field width for noncharacter objects.NULL
corresponds to the default12
. na.encode
 logical: should
NA
strings be encoded? Note this only applies to elements of character vectors, not to numerical, complex nor logicalNA
s, which are always encoded as"NA"
.  scientific
 Either a logical specifying whether
elements of a real or complex vector should be encoded in scientific
format, or an integer penalty (see
options("scipen")
). Missing values correspond to the current default penalty.  …
 further arguments passed to or from other methods.
 big.mark, big.interval, small.mark, small.interval, decimal.mark, zero.print, drop0trailing

used for prettying (longish) numerical and complex sequences.
Passed to
prettyNum
: that help page explains the details.
Details
format
is a generic function. Apart from the methods described
here there are methods for dates (see format.Date
),
datetimes (see format.POSIXct
)) and for other classes such
as format.octmode
and format.dist
. format.data.frame
formats the data frame column by column,
applying the appropriate method of format
for each column.
Methods for columns are often similar to as.character
but offer
more control. Matrix and dataframe columns will be converted to
separate columns in the result, and character columns (normally all)
will be given class "AsIs"
. format.factor
converts the factor to a character vector and
then calls the default method (and so justify
applies). format.AsIs
deals with columns of complicated objects that
have been extracted from a data frame. Character objects and (atomic)
matrices are passed to the default method (and so width
does
not apply).
Otherwise it calls toString
to convert the object
to character (if a vector or list, element by element) and then
rightjustifies the result. Justification for character vectors (and objects converted to
character vectors by their methods) is done on display width (see
nchar
), taking doublewidth characters and the rendering
of special characters (as escape sequences, including escaping
backslash but not double quote: see print.default
) into
account. Thus the width is as displayed by print(quote =
FALSE)
and not as displayed by cat
. Character strings
are padded with blanks to the display width of the widest. (If
na.encode = FALSE
missing character strings are not included in
the width computations and are not encoded.) Numeric vectors are encoded with the minimum number of decimal places
needed to display all the elements to at least the digits
significant digits. However, if all the elements then have trailing
zeroes, the number of decimal places is reduced until
nsmall
is reached or at least one
element has a nonzero final digit; see also the argument
documentation for big.*
, small.*
etc, above. See the
note in print.default
about digits >= 16
. Raw vectors are converted to their 2digit hexadecimal representation
by as.character
. The internal code respects the option
getOption("OutDec")
for the ‘decimal mark’, so if
this is set to something other than "."
then it takes precedence
over argument decimal.mark
.
Value
An object of similar structure to x
containing character
representations of the elements of the first argument x
in a common format, and in the current locale's encoding. For character, numeric, complex or factor x
, dims and dimnames
are preserved on matrices/arrays and names on vectors: no other
attributes are copied. If x
is a list, the result is a character vector obtained by
applying format.default(x, …)
to each element of the list
(after unlist
ing elements which are themselves lists),
and then collapsing the result for each element with
paste(collapse = ", ")
. The defaults in this case are
trim = TRUE, justify = "none"
since one does not usually want
alignment in the collapsed strings.
References
Becker, R. A., Chambers, J. M. and Wilks, A. R. (1988) The New S Language. Wadsworth & Brooks/Cole.
See Also
format.info
indicates how an atomic vector would be
formatted. formatC
, paste
, as.character
,
sprintf
, print
, prettyNum
,
toString
, encodeString
.
Examples
library(base)
format(1:10)
format(1:10, trim = TRUE)
zz < data.frame("(row names)"= c("aaaaa", "b"), check.names = FALSE)
format(zz)
format(zz, justify = "left")
## use of nsmall
format(13.7)
format(13.7, nsmall = 3)
format(c(6.0, 13.1), digits = 2)
format(c(6.0, 13.1), digits = 2, nsmall = 1)
## use of scientific
format(2^311)
format(2^311, scientific = TRUE)
## a list
z < list(a = letters[1:3], b = (pi+0i)^((2:2)/2), c = c(1,10,100,1000),
d = c("a", "longer", "character", "string"),
q = quote( a + b ), e = expression(1+x))
## can you find the "2" small differences?
(f1 < format(z, digits = 2))
(f2 < format(z, digits = 2, justify = "left", trim = FALSE))
f1 == f2 ## 2 FALSE, 4 TRUE
Community examples
By default, `format()` adds leading spaces so that all values have the same length. Setting `trim = TRUE` turns this off. ```{r} x < 10 ^ (0:4) format(x) format(x, trim = TRUE) ``` `digits` sets the number of significant digits. ```{r} x2 < c(0, 0.1, 0.12, 0.123, 0.1234) format(x2, digits = 3) ``` Actually, it sets significant digits for the smallest (in magnitude) value, so if you have orders of magnitude range, treat it as a suggestion and fiddle as necessary. ```{r} x < 10 ^ (0:4) x2 < c(0, 0.1, 0.12, 0.123, 0.1234) format(x + x2, digits = 3) ``` `digits` can also be useful for examining floating point rounding errors. ```{r} format(sqrt(2) * sqrt(2)  2, digits = 22) ``` `nsmall` gives the minimum number of decimal places. ```{r} x < 10 ^ (0:4) x2 < c(0, 0.1, 0.12, 0.123, 0.1234) format(x + x2, nsmall = 3) ``` You can use `digits` and `nsmall` together; the result has enough digits to satisfy the sig digs and the d.p. criteria. ```{r} x < 10 ^ (0:4) x2 < c(0, 0.1, 0.12, 0.123, 0.1234) format(x + x2, digits = 5, nsmall = 1) format(x + x2, digits = 1, nsmall = 5) ``` `justify` lets you enforce left/right/centre justification on character vectors. It has no effect when the input is a number. ```{r} x < 10 ^ (0:4) format(x, justify = "centre") # does nothing format(format(x, trim = TRUE), justify = "left") format(format(x, trim = TRUE), justify = "centre") ``` `width` pads values with leading spaces to set a minimum output width. ```{r} format(0, width = 10) format(1.2345, width = 1) # width is greater than 1 anyway ``` For objects of class `AsIs`, width sets the maximum output width. ```{r} x < 10 ^ (0:4) format(I(x), width = 3) ``` Usually `format()` turns `NA` into `"NA"`. Setting `na.encode = FALSE` turns this behaviour off for character vectors. ```{r} format(c("abc", NA), na.encode = FALSE) format(c(1, NA), na.encode = FALSE) # no effect ``` For numeric vectors, you can turn scientific formatting on or off iwth the `scientific` argument. ```{r} format(2 ^ 31  1, scientific = FALSE) format(2 ^ 31  1, scientific = TRUE) ``` This doesn't work for integers. ```{r} format(.Machine$integer.max, scientific = TRUE) ``` There are more arguments for customising the output for written reports. See [`prettyNum()`](https://www.rdocumentation.org/packages/base/topics/prettyNum) for more. ```{r} format( 123456789.87654321, digits = 17, big.mark = ",", small.mark = " ", small.interval = 2 ) ``` For data frames, the arguments passed to format work on all columns. ```{r} co2 < CO2[sample(nrow(CO2), 12), ] format(co2, justify = "left", width = 10, scientific = TRUE) ``` Factors default to left justification (in a lefttoright locale?) ```{r} f < factor(sample(month.name, 25, replace = TRUE)) format(f) format(f, justify = "right", width = 12) ```