Numbers¶

Standard Numeric Types¶

BoolInt8UInt8Int16UInt16Int32UInt32Int64UInt64Int128UInt128Float16Float32Float64Complex64Complex128

Data Formats¶

bin(n[, pad])¶: Convert an integer to a binary string, optionally specifying a number of digits to pad to.

hex(n[, pad])¶: Convert an integer to a hexadecimal string, optionally specifying a number of digits to pad to.

dec(n[, pad])¶: Convert an integer to a decimal string, optionally specifying a number of digits to pad to.

oct(n[, pad])¶: Convert an integer to an octal string, optionally specifying a number of digits to pad to.

base(base, n[, pad])¶: Convert an integer to a string in the given base, optionally specifying a number of digits to pad to.

digits([T, ]n[, base][, pad])¶: Returns an array with element type T (default Int) of the digits of n in the given base, optionally padded with zeros to a specified size. More significant digits are at higher indexes, such that n==sum([digits[k]*base^(k-1)fork=1:length(digits)]).

digits!(array, n[, base])¶: Fills an array of the digits of n in the given base. More significant digits are at higher indexes. If the array length is insufficient, the least significant digits are filled up to the array length. If the array length is excessive, the excess portion is filled with zeros.

bits(n)¶: A string giving the literal bit representation of a number.

parse(type, str[, base])¶: Parse a string as a number. If the type is an integer type, then a base can be specified (the default is 10). If the type is a floating point type, the string is parsed as a decimal floating point number. If the string does not contain a valid number, an error is raised.

tryparse(type, str[, base])¶: Like parse, but returns a Nullable of the requested type. The result will be null if the string does not contain a valid number.

big(x)¶: Convert a number to a maximum precision representation (typically BigInt or BigFloat). See BigFloat for information about some pitfalls with floating-point numbers.

signed(x)¶: Convert a number to a signed integer. If the argument is unsigned, it is reinterpreted as signed without checking for overflow.

unsigned(x) → Unsigned¶: Convert a number to an unsigned integer. If the argument is signed, it is reinterpreted as unsigned without checking for negative values.

float(x)¶: Convert a number, array, or string to a AbstractFloat data type. For numeric data, the smallest suitable AbstractFloat type is used. Converts strings to Float64.

significand(x)¶

Extract the significand(s) (a.k.a. mantissa), in binary representation, of a floating-point number or array. If x is a non-zero finite number, then the result will be a number of the same type on the interval \([1,2)\). Otherwise x is returned.

julia>significand(15.2)/15.20.125julia>significand(15.2)*815.2

exponent(x) → Int¶: Get the exponent of a normalized floating-point number.

complex(r[, i])¶: Convert real numbers or arrays to complex. i defaults to zero.

bswap(n)¶: Byte-swap an integer.

num2hex(f)¶: Get a hexadecimal string of the binary representation of a floating point number.

hex2num(str)¶: Convert a hexadecimal string to the floating point number it represents.

hex2bytes(s::AbstractString)¶: Convert an arbitrarily long hexadecimal string to its binary representation. Returns an Array{UInt8,1}, i.e. an array of bytes.

bytes2hex(bin_arr::Array{UInt8, 1})¶: Convert an array of bytes to its hexadecimal representation. All characters are in lower-case. Returns a String.

General Number Functions and Constants¶

one(x)¶: Get the multiplicative identity element for the type of x (x can also specify the type itself). For matrices, returns an identity matrix of the appropriate size and type.

zero(x)¶: Get the additive identity element for the type of x (x can also specify the type itself).

pi¶
π¶: The constant pi.

im¶: The imaginary unit.

e¶
eu¶: The constant e.

catalan¶: Catalan’s constant.

γ¶
eulergamma¶: Euler’s constant.

φ¶
golden¶: The golden ratio.

Inf¶: Positive infinity of type Float64.

Inf32¶: Positive infinity of type Float32.

Inf16¶: Positive infinity of type Float16.

NaN¶: A not-a-number value of type Float64.

NaN32¶: A not-a-number value of type Float32.

NaN16¶: A not-a-number value of type Float16.

issubnormal(f) → Bool¶: Test whether a floating point number is subnormal.

isfinite(f) → Bool¶: Test whether a number is finite

isinf(f) → Bool¶: Test whether a number is infinite.

isnan(f) → Bool¶: Test whether a floating point number is not a number (NaN).

nextfloat(x::AbstractFloat)¶: Returns the smallest floating point number y of the same type as x such x<y. If no such y exists (e.g. if x is Inf or NaN), then returns x.

prevfloat(x::AbstractFloat)¶: Returns the largest floating point number y of the same type as x such y<x. If no such y exists (e.g. if x is -Inf or NaN), then returns x.

nextfloat(x::AbstractFloat, n::Integer): The result of n iterative applications of nextfloat to x if n>=0, or -n applications of prevfloat if n<0.

isinteger(x) → Bool¶: Test whether x or all its elements are numerically equal to some integer

isreal(x) → Bool¶: Test whether x or all its elements are numerically equal to some real number.

isimag(z) → Bool¶: Test whether z is purely imaginary, i.e. has a real part equal to 0.

Float32(x[, mode::RoundingMode])¶

Create a Float32 from x. If x is not exactly representable then mode determines how x is rounded.

julia>Float32(1/3,RoundDown)0.3333333f0julia>Float32(1/3,RoundUp)0.33333334f0

See RoundingMode for available rounding modes.

Float64(x[, mode::RoundingMode])¶

Create a Float64 from x. If x is not exactly representable then mode determines how x is rounded.

julia>Float64(pi,RoundDown)3.141592653589793julia>Float64(pi,RoundUp)3.1415926535897936

See RoundingMode for available rounding modes.

BigInt(x)¶

Create an arbitrary precision integer. x may be an Int (or anything that can be converted to an Int). The usual mathematical operators are defined for this type, and results are promoted to a BigInt.

Instances can be constructed from strings via parse(), or using the big string literal.

BigFloat(x)¶

Create an arbitrary precision floating point number. x may be an Integer, a Float64 or a BigInt. The usual mathematical operators are defined for this type, and results are promoted to a BigFloat.

Note that because decimal literals are converted to floating point numbers when parsed, BigFloat(2.1) may not yield what you expect. You may instead prefer to initialize constants from strings via parse(), or using the big string literal.

julia>BigFloat(2.1)2.100000000000000088817841970012523233890533447265625000000000000000000000000000julia>big"2.1"2.099999999999999999999999999999999999999999999999999999999999999999999999999986

rounding(T)¶

Get the current floating point rounding mode for type T, controlling the rounding of basic arithmetic functions (+(), -(), *(), /() and sqrt()) and type conversion.

See RoundingMode for available modes.

setrounding(T, mode)¶

Set the rounding mode of floating point type T, controlling the rounding of basic arithmetic functions (+(), -(), *(), /() and sqrt()) and type conversion. Other numerical functions may give incorrect or invalid values when using rounding modes other than the default RoundNearest.

Note that this may affect other types, for instance changing the rounding mode of Float64 will change the rounding mode of Float32. See RoundingMode for available modes.

Warning

This feature is still experimental, and may give unexpected or incorrect values.

setrounding(f::Function, T, mode)

Change the rounding mode of floating point type T for the duration of f. It is logically equivalent to:

old=rounding(T)setrounding(T,mode)f()setrounding(T,old)

See RoundingMode for available rounding modes.

Warning

This feature is still experimental, and may give unexpected or incorrect values. A known problem is the interaction with compiler optimisations, e.g.

julia>setrounding(Float64,RoundDown)do1.1+0.1end1.2000000000000002

Here the compiler is constant folding, that is evaluating a known constant expression at compile time, however the rounding mode is only changed at runtime, so this is not reflected in the function result. This can be avoided by moving constants outside the expression, e.g.

julia>x=1.1;y=0.1;julia>setrounding(Float64,RoundDown)dox+yend1.2

get_zero_subnormals() → Bool¶: Returns false if operations on subnormal floating-point values (“denormals”) obey rules for IEEE arithmetic, and true if they might be converted to zeros.

set_zero_subnormals(yes::Bool) → Bool¶

If yes is false, subsequent floating-point operations follow rules for IEEE arithmetic on subnormal values (“denormals”). Otherwise, floating-point operations are permitted (but not required) to convert subnormal inputs or outputs to zero. Returns true unless yes==true but the hardware does not support zeroing of subnormal numbers.

set_zero_subnormals(true) can speed up some computations on some hardware. However, it can break identities such as (x-y==0)==(x==y).

Integers¶

count_ones(x::Integer) → Integer¶

Number of ones in the binary representation of x.

julia>count_ones(7)3

count_zeros(x::Integer) → Integer¶

Number of zeros in the binary representation of x.

julia>count_zeros(Int32(2^16-1))16

leading_zeros(x::Integer) → Integer¶

Number of zeros leading the binary representation of x.

julia>leading_zeros(Int32(1))31

leading_ones(x::Integer) → Integer¶

Number of ones leading the binary representation of x.

julia>leading_ones(UInt32(2^32-2))31

trailing_zeros(x::Integer) → Integer¶

Number of zeros trailing the binary representation of x.

julia>trailing_zeros(2)1

trailing_ones(x::Integer) → Integer¶

Number of ones trailing the binary representation of x.

julia>trailing_ones(3)2

isodd(x::Integer) → Bool¶

Returns true if x is odd (that is, not divisible by 2), and false otherwise.

julia>isodd(9)truejulia>isodd(10)false

iseven(x::Integer) → Bool¶

Returns true is x is even (that is, divisible by 2), and false otherwise.

julia>iseven(9)falsejulia>iseven(10)true

BigFloats¶

The BigFloat type implements arbitrary-precision floating-point arithmetic using the GNU MPFR library.

precision(num::AbstractFloat)¶: Get the precision of a floating point number, as defined by the effective number of bits in the mantissa.

precision(BigFloat): Get the precision (in bits) currently used for BigFloat arithmetic.

setprecision([T=BigFloat, ]precision::Int)¶: Set the precision (in bits) to be used for T arithmetic.

setprecision(f::Function, [T=BigFloat, ]precision::Integer)

Change the T arithmetic precision (in bits) for the duration of f. It is logically equivalent to:

old=precision(BigFloat)setprecision(BigFloat,precision)f()setprecision(BigFloat,old)

Often used as setprecision(T,precision)do...end

Random Numbers¶

Random number generation in Julia uses the Mersenne Twister library via MersenneTwister objects. Julia has a global RNG, which is used by default. Other RNG types can be plugged in by inheriting the AbstractRNG type; they can then be used to have multiple streams of random numbers. Besides MersenneTwister, Julia also provides the RandomDevice RNG type, which is a wrapper over the OS provided entropy.

Most functions related to random generation accept an optional AbstractRNG as the first argument, rng , which defaults to the global one if not provided. Morever, some of them accept optionally dimension specifications dims... (which can be given as a tuple) to generate arrays of random values.

A MersenneTwister or RandomDevice RNG can generate random numbers of the following types: Float16, Float32, Float64, Bool, Int8, UInt8, Int16, UInt16, Int32, UInt32, Int64, UInt64, Int128, UInt128, BigInt (or complex numbers of those types). Random floating point numbers are generated uniformly in \([0, 1)\). As BigInt represents unbounded integers, the interval must be specified (e.g. rand(big(1:6))).

srand([rng][, seed])¶: Reseed the random number generator. If a seed is provided, the RNG will give a reproducible sequence of numbers, otherwise Julia will get entropy from the system. For MersenneTwister, the seed may be a non-negative integer, a vector of UInt32 integers or a filename, in which case the seed is read from a file. RandomDevice does not support seeding.

MersenneTwister([seed])¶: Create a MersenneTwister RNG object. Different RNG objects can have their own seeds, which may be useful for generating different streams of random numbers.

RandomDevice()¶: Create a RandomDevice RNG object. Two such objects will always generate different streams of random numbers.

rand([rng][, S][, dims...])¶

Pick a random element or array of random elements from the set of values specified by S; S can be

an indexable collection (for example 1:n or ['x','y','z']), or
a type: the set of values to pick from is then equivalent to typemin(S):typemax(S) for integers (this is not applicable to BigInt), and to \([0, 1)\) for floating point numbers;

S defaults to Float64.

rand!([rng, ]A[, coll])¶: Populate the array A with random values. If the indexable collection coll is specified, the values are picked randomly from coll. This is equivalent to copy!(A,rand(rng,coll,size(A))) or copy!(A,rand(rng,eltype(A),size(A))) but without allocating a new array.

bitrand([rng][, dims...])¶: Generate a BitArray of random boolean values.

randn([rng][, T=Float64][, dims...])¶: Generate a normally-distributed random number of type T with mean 0 and standard deviation 1. Optionally generate an array of normally-distributed random numbers. The Base module currently provides an implementation for the types Float16, Float32, and Float64 (the default).

randn!([rng, ]A::AbstractArray) → A¶: Fill the array A with normally-distributed (mean 0, standard deviation 1) random numbers. Also see the rand function.

randexp([rng][, T=Float64][, dims...])¶: Generate a random number of type T according to the exponential distribution with scale 1. Optionally generate an array of such random numbers. The Base module currently provides an implementation for the types Float16, Float32, and Float64 (the default).

randexp!([rng, ]A::AbstractArray) → A¶: Fill the array A with random numbers following the exponential distribution (with scale 1).

randjump(r::MersenneTwister, jumps[, jumppoly]) → Vector{MersenneTwister}¶

Create an array of the size jumps of initialized MersenneTwister RNG objects where the first RNG object given as a parameter and following MersenneTwister RNGs in the array initialized such that a state of the RNG object in the array would be moved forward (without generating numbers) from a previous RNG object array element on a particular number of steps encoded by the jump polynomial jumppoly.

Default jump polynomial moves forward MersenneTwister RNG state by 10^20 steps.