Julia Data Types
While Julia is a dynamically typed language, values have a concrete type.
Contents
Abstract Types
Abstract types should not be used directly. They are supertypes that categorize behaviors. They are however useful for constructing a novel data type, or for typing the arguments to a function.
The primary set of abstract types, with their hierarchy of supertypes and subtypes indicated by indentation, is:
Number
Real
AbstractFloat
Integer
Signed
Unsigned
AbstractChar
AbstractString
AbstractArray
AbstractVector
AbstractMatrix
AbstractRange
Numeric Types
The set of primitive types that descend from the Number abstract type are:
Bool
Int8
Int16
Int32
Int64
Int128
UInt8
UInt16
UInt32
UInt64
UInt128
Float16
Float32
Float64
String Types
String is a subtype of AbstractString that implicitly is encoded in UTF-8.
Packages may create novel subtypes of AbstractString, so consider typing string arguments for functions as AbstractString rather than String.
Char is a subtype of AbstractChar, and is a 32-bit representation a Unicode character.
Composite Types
Structs are a composite data type. Struct values can be instantiated like:
julia> struct Foo
bar
baz
qux
end
julia> foo = Foo("Hello, world.", 23, 1.5)
Foo("Hello, world.", 23, 1.5)Fields are accessed by the dot operator.
julia> foo.bar "Hello, world."
By default, struct objects are immutable.
Collection Types
Values can be collected in a variety of series types.
The Array type is a general purpose series of values, all of the same type. Both Vector and Matrix types descend from AbstractArray.
The range operator (:) creates collections that have some type descending from AbstractRange, which also descends from AbstractArray. The differences between e.g. UnitRange and OrdinalRange relate to the available optimizations.
== Nothing Type ===
The singleton constant nothing is of type Nothing.
