Learn R Programming

gdata (version 2.19.0)

humanReadable: Print Byte Size in Human Readable Format

Description

humanReadable converts integer byte sizes to a human readable units such as kB, MB, GB, etc.

Usage

humanReadable(x, units="auto", standard=c("IEC", "SI", "Unix"),
              digits=1, width=NULL, sep=" ", justify=c("right", "left") )

Value

Byte size in human readable format as character with proper unit symbols added at the end of the string.

Arguments

x

integer, byte size

standard

character, "IEC" for powers of 1024 ('MiB'), "SI" for powers of 1000 ('MB'), or "Unix" for powers of 1024 ('M'). See details.

units

character, unit to use for all values (optional), one of "auto", "bytes", or an appropriate unit corresponding to standard.

digits

integer, number of digits after decimal point

width

integer, width of number string

sep

character, separator between number and unit

justify

two-element vector specifiy the alignment for the number and unit components of the size. Each element should be one of "none", "left", "right", or "center"

Author

Ales Korosec, Gregor Gorjanc, and Gregory R. Warnes greg@warnes.net

Details

The basic unit used to store information in computers is a bit. Bits are represented as zeroes and ones - binary number system. Although, the binary number system is not the same as the decimal number system, decimal prefixes for binary multiples such as kilo and mega are often used. In the decimal system kilo represent 1000, which is close to \(1024 = 2^{10}\) in the binary system. This sometimes causes problems as it is not clear which powers (2 or 10) are used in a notation like 1 kB. To overcome this problem International Electrotechnical Commission (IEC) has provided the following solution to this problem:

NameSystemSymbolSizeConversion
bytebinaryB\(2^3\)8 bits
kilobytedecimalkB\(10^3\)1000 bytes
kibibytebinaryKiB\(2^{10}\)1024 bytes
megabytedecimalMB\((10^3)^2\)1000 kilobytes
mebibytebinaryMiB\((2^{10})^2\)1024 kibibytes
gigabytedecimalGB\((10^3)^3\)1000 megabytes
gibibytebinaryGiB\((2^{10})^3\)1024 mebibytes
terabytedecimalTB\((10^3)^4\)1000 gigabytes
tebibytebinaryTiB\((2^{10})^4\)1024 gibibytes
petabytedecimalPB\((10^3)^5\)1000 terabytes
pebibytebinaryPiB\((2^{10})^5\)1024 tebibytes
exabytedecimalEB\((10^3)^6\)1000 petabytes
exbibytebinaryEiB\((2^{10})^6\)1024 pebibytes
zettabytedecimalZB\((10^3)^7\)1000 exabytes
zebibytebinaryZiB\((2^{10})^7\)1024 exbibytes
yottabytedecimalYB\((10^3)^8\)1000 zettabytes
yebibytebinaryYiB\((2^{10})^8\)1024 zebibytes

where Zi and Yi are GNU extensions to IEC. To get the output in the decimal system (powers of 1000) use standard="SI". To obtain IEC standard (powers of 1024) use standard="IEC".

In addition, single-character units are provided that follow (and extend) the Unix pattern (use standard="Unix"):

NameSystemSymbolSizeConversion
bytebinaryB\(2^3\)8 bits
kibibytebinaryK\(2^{10}\)1024 bytes
mebibytebinaryM\((2^{10})^2\)1024 kibibytes
gibibytebinaryG\((2^{10})^3\)1024 mebibytes
tebibytebinaryT\((2^{10})^4\)1024 gibibytes
pebibytebinaryP\((2^{10})^5\)1024 tebibytes
exbibytebinaryE\((2^{10})^6\)1024 pebibytes
zebibytebinaryZ\((2^{10})^7\)1024 exbibytes
yottabytebinaryY\((2^{10})^8\)1024 zebibytes

For printout both digits and width can be specified. If width is NULL, all values have given number of digits. If width is not NULL, output is rounded to a given width and formated similar to human readable format of the Unix ls, df or du shell commands.

References

Wikipedia: https://en.wikipedia.org/wiki/Byte https://en.wikipedia.org/wiki/SI_prefix https://en.wikipedia.org/wiki/Binary_prefix

See Also

object.size in package 'gdata', object.size in package 'utils', ll

Examples

Run this code

# Simple example: maximum addressible size of 32 bit pointer
humanReadable(2^32-1)
humanReadable(2^32-1, standard="IEC")
humanReadable(2^32-1, standard="SI")
humanReadable(2^32-1, standard="Unix")

humanReadable(2^32-1, unit="MiB")
humanReadable(2^32-1, standard="IEC",  unit="MiB")
humanReadable(2^32-1, standard="SI",   unit="MB")
humanReadable(2^32-1, standard="Unix", unit="M")

# Vector of sizes
matrix(humanReadable(c(60810, 124141, 124, 13412513), width=4))
matrix(humanReadable(c(60810, 124141, 124, 13412513), width=4, unit="KiB"))

# Specify digits rather than width
matrix(humanReadable(c(60810, 124141, 124, 13412513), width=NULL, digits=2))

# Change the justification
matrix(humanReadable(c(60810, 124141, 124, 13412513), width=NULL,
                    justify=c("right", "right") ))

Run the code above in your browser using DataLab