MATLAB to Python Data Type Mapping

When calling a Python^® function, MATLAB^® converts MATLAB data into types that best represent the data to the Python language.

Pass Scalar Values to Python

MATLAB Input Argument Type — Scalar Values Only	Resulting Python `py.` Type	Examples
`double` (real) `single` (real)	`float`	Use Python Numeric Variables in MATLAB
`double` (complex) `single` (complex)	`complex`	z = complex(1,2); py.cmath.polar(z) ans = Python tuple with no properties. (2.23606797749979, 1.1071487177940904)
`int8` `uint8` `int16` `uint16` `int32`	`int`
`uint32` `int64` `uint64`	`int` `long` (version 2.7 only)
`NaN`	`float("nan")`
`Inf`	`float("inf")`
`string` scalar `char` vector	`str`	Use Python str Variables in MATLAB
`<missing>` value in `string`	`None`	py.list({string(missing),'Value'}) ans = Python list with no properties. [None, 'Value']
`logical`	`bool`
Structure	`dict`	Use Python dict Variables in MATLAB
Python object — `py`.`type`	`type`
function handle `@py`.`module`.`function`, to Python functions only	`module`.`function`	Pass Python Function to Python map Function

Pass Vectors to Python

MATLAB Input Argument Type — `1`-by-`N` Vector	Resulting Python Type
`double` (real)	`array.array('d')`
`single` (real)	`array.array('f')`
`int8` (real)	`array.array('b')`
`uint8` (real)	`array.array('B')`
`int16` (real)	`array.array('h')`
`uint16` (real)	`array.array('H')`
`int32` (real)	`array.array('i')`
`uint32` (real)	`array.array('I')`
`int64` (real) - Not supported for Python 2.7 on Windows^®	`array.array('q')`
`uint64` (real) - Not supported for Python 2.7 on Windows	`array.array('Q')`
`double` (complex) `single` (complex) `int8` (complex) `uint8` (complex) `int16` (complex) `uint16` (complex) `int32` (complex) `uint32` (complex)	`memoryview`
`logical`	`memoryview`
`char` vector `string` scalar	`str`
`char` array containing values greater than 127 (version 2.7 only)	`unicode`
`cell` vector	`tuple`

Pass Matrices and Multidimensional Arrays to Python

The Python language provides a protocol for accessing memory buffers like the data stored in a MATLAB array. MATLAB implements this Python buffer protocol for MATLAB arrays so that you can read MATLAB arrays directly from Python code, running in the same process as MATLAB, without copying data.

Many Python functions directly use the MATLAB array from Python without converting it to a native Python type. Some functions might require a specific type, such as numpy.ndarray, or might modify data in the array. These functions might accept the MATLAB array and copy the data into the required type. Other functions might display an error if you do not pass the required type. To pass data to these functions, first create the required Python type from the MATLAB data, then pass it to the Python function. For example, to create array p to pass to a Python function that requires data of type numpy.array, type:

p = py.numpy.array(magic(3))

p = 

  Python ndarray:

     8     1     6
     3     5     7
     4     9     2

    Use details function to view the properties of the Python object.

    Use double function to convert to a MATLAB array.

MATLAB sparse arrays are not supported in Python. See Unsupported MATLAB Types.

Troubleshooting Argument Errors

If a Python function expects a specific Python multidimensional array type such as numpy.ndarray, then MATLAB displays a message with tips about how to proceed. If the problem might be due to passing a matrix or a multidimensional array as an argument, then do the following.

Check the documentation for the Python function and find out the expected type for the argument.
Create a Python object of that type in MATLAB and pass that to the Python function.

For example, suppose that the following code returns an error.

a = [1 2; 3 4];
py.pyfunc(a)

If the documentation of pyfunc specifies that the expected type is numpy.ndarray, then try this conversion:

py.pyfunc(numpy.ndarray(a))

If the error persists, then determine the root cause by checking for additional information in the Python exception.

Automatically Convert Python Types to MATLAB Types

MATLAB automatically converts these data types returned from Python functions into MATLAB types. To convert other types, see Explicitly Convert Python Types to MATLAB Types.

Python Return Type, as Displayed in Python	Resulting MATLAB Type — Scalar
`float`	`double`
`complex`	Complex `double`
`int` (version 2.7 only). For Python versions 3.x `int`, you must convert explicitly. See Explicitly Convert Python Types to MATLAB Types.	`int64`
`bool`	`logical`
All other Python types — `type`	Python object — `py`.`type`

Explicitly Convert Python Types to MATLAB Types

If the output of a Python function implements the Python buffer protocol, such as numpy.ndarray, and it is numeric or logical, then MATLAB displays:

The actual Python type
The underlying data
The corresponding MATLAB conversion function. Use this function to fully convert the Python object to a MATLAB array.

Use these MATLAB functions to convert Python data types to MATLAB types.

Python Return Type or Protocol, as Displayed in MATLAB	MATLAB Conversion Function	Examples
`py.str` (version 3.x)	`string` `char`	Use Python str Variables in MATLAB
`py.str` (version 2.7)	`string` `char` `uint8`
`py.unicode`	`string` `char`
Object with `__str__` method	`char`	py.help('datetime.date.__str__') Help on wrapper_descriptor in datetime.date: datetime.date.__str__ = __str__(self, /) Return str(self). d = py.datetime.date(... int32(2020),int32(3),int32(4)); char(d) ans = '2020-3-04'
`py.bytes`	`uint8`
`py.int`	`double` or `int64`
`py.long`	`double` or `int64`
`py.array.array` `py.numpy.ndarray` `py.memoryview` You can convert `py.array.array` of any format and `py.memoryview` objects to the MATLAB type you want.	Numeric functions: `double` `single` `int8` `uint8` `int16` `uint16` `int32` `uint32` `int64` `uint64`	Use Python Numeric Variables in MATLAB, for example, Use Python Integer array Types in MATLAB.
`py.list` and `py.tuple`	`double` `single` `int8` `uint8` `int16` `uint16` `int32` `uint32` `int64` `uint64` `logical` `string` `cell`	Use Python list Variables in MATLAB Use Python tuple Variables in MATLAB For more information, see Error Converting Elements of list or tuple.
Mapping protocol; for example, `py.dict`	`struct`	Use Python dict Variables in MATLAB

For example, a Python function returns this array p:

p = 

  Python ndarray:

     8     1     6
     3     5     7
     4     9     2

    Use details function to view the properties of the Python object.

    Use double function to convert to a MATLAB array.

You can convert it to a MATLAB matrix P by typing:

P = double(p)

P = 3×3    
     8     1     6
     3     5     7
     4     9     2

If you need specific information about the Python properties of p, type:

details(p)

    py.numpy.ndarray handle with properties:

           T: [1×1 py.numpy.ndarray]
        base: [1×1 py.NoneType]
      ctypes: [1×1 py.numpy.core._internal._ctypes]
        data: [1×3 py.memoryview]
       dtype: [1×1 py.numpy.dtype[float64]]
       flags: [1×1 py.numpy.flagsobj]
        flat: [1×1 py.numpy.flatiter]
        imag: [1×1 py.numpy.ndarray]
    itemsize: [1×1 py.int]
      nbytes: [1×1 py.int]
        ndim: [1×1 py.int]
        real: [1×1 py.numpy.ndarray]
       shape: [1×2 py.tuple]
        size: [1×1 py.int]
     strides: [1×2 py.tuple]

  Methods, Events, Superclasses

If the Python module provides content in its __doc__ attribute, then MATLAB links to that information.

Unsupported MATLAB Types

These MATLAB types are not supported in Python.

Multidimensional char or cell arrays
Structure arrays
Sparse arrays
categorical,
table,
containers.Map,
datetime types
MATLAB objects
meta.class (py.class)