Matlab Types

• MATLAB stores and accesses all data through the array structure; “data type” refers to the type of the elements of an array
• MATLAB’s data types are described as classes with attributes
  • For example, a complex number has class double with attribute complex
  • As with Python Sequence in Python, you can index, slice, and so on
• There are [15]“R2012b”+x built-in fundamental types and 2 user-defined kinds^[https://www.mathworks.com/help/matlab/matlab_prog/fundamental-matlab-classes.html]
• Fundamental data types:
  • Numeric
  • Character
  • Structure
  • Cell
  • Logical
  • Function Handle
• Type determination functions (returning logical values)
  • Matlab Functions - class
  • isa(A, 'class_name')
  • isreal(A)
  • isnan(A)
  • isnumeric(A)
  • isinf(A)
  • isinteger(A)
  • isfinite(A)
  • isfloat(A)

graph TD
A{{Array}} --> B(Numeric)
    B --> B1(Float)
        B1 --> B11(Single)
        B1 --> B12(Double):::imp
    B --> B2(Integer)
        B2 --> B21(Signed)
            B21 --- B211(int8-bit)
            B211 --- B212(int16-bit)
            B212 --- B213(int32-bit)
            B213 --- B214(int64-bit)
        B2 --> B22(Unsigned)
            B22 --- B221(uint8-bit)
            B221 --- B222(uint16-bit)
            B222 --- B223(uint32-bit)
            B223 --- B224(uint64-bit)
A --> C(Text)
    C --> C1(Character):::imp
    C --> C2(String)
A --> D(Function Handle):::imp
A --> E(Heterogeneous Container)
    E --> E1(Cell):::imp
    E --> E2(Structure):::imp
A --> F(Logical):::imp

classDef imp stroke:#ff0000,stroke-width:4px;

• The string type was introduced later; the difference from character can be roughly read from the names — see Matlab Characters and Strings
  • A character is analogous to a single value within a numeric array; each character occupies exactly 2 bytes
  • A string behaves more like a numeric array taken as a single value
• Integer types come in signed and unsigned flavours; the former includes negative values, the latter does not
• The trailing number indicates the bit width of the binary representation
  • For the same width, unsigned covers a larger non-negative range than signed, since one bit is reserved for the sign

Types Conversion

A