2.6. Supported Python features¶

2.6.1.1. Constructs¶

Numba strives to support as much of the Python language as possible, but some language features are not available inside Numba-compiled functions. The following Python language features are not currently supported:

• Class definition
• Exception handling (try .. except, try .. finally)
• Context management (the with statement)
• Some comprehensions (list comprehension is supported, but not dict, set or generator comprehensions)
• Generator delegation (yield from)

The raise statement is supported in several forms:

• raise (to re-raise the current exception)
• raise SomeException
• raise SomeException(<arguments>): in nopython mode, constructor arguments must be compile-time constants

Similarly, the assert statement is supported with or without an error message.

2.6.1.2. Functions¶

from numba import jit

@jit
def add1(x):
    return x + 1

@jit
def bar(fn, x):
    return fn(x)

@jit
def foo(x):
    return bar(add1, x)

# Passing add1 within numba compiled code.
print(foo(1))
# Passing add1 into bar from interpreted code
print(bar(add1, 1))

2.6.1.3. Generators¶

Numba supports generator functions and is able to compile them in object mode and nopython mode. The returned generator can be used both from Numba-compiled code and from regular Python code.

Coroutine features of generators are not supported (i.e. the generator.send(), generator.throw(), generator.close() methods).

2.6.2.1. int, bool¶

Arithmetic operations as well as truth values are supported.

The following attributes and methods are supported:

• .conjugate()
• .real
• .imag

2.6.2.2. float, complex¶

Arithmetic operations as well as truth values are supported.

The following attributes and methods are supported:

• .conjugate()
• .real
• .imag

2.6.2.3. str¶

Numba supports (Unicode) strings in Python 3. Strings can be passed into nopython mode as arguments, as well as constructed and returned from nopython mode. As in Python, slices (even of length 1) return a new, reference counted string. Optimized code paths for efficiently accessing single characters may be introduced in the future.

The in-memory representation is the same as was introduced in Python 3.4, with each string having a tag to indicate whether the string is using a 1, 2, or 4 byte character width in memory. When strings of different encodings are combined (as in concatenation), the resulting string automatically uses the larger character width of the two input strings. String slices also use the same character width as the original string, even if the slice could be represented with a narrower character width. (These details are invisible to the user, of course.)

The following functions, attributes and methods are currently supported:

• len()
• + (concatenation of strings)
• in, .contains()
• ==, <, <=, >, >= (comparison)
• .startswith()
• .endswith()
• .find()

Additional operations as well as support for Python 2 strings / Python 3 bytes will be added in a future version of Numba. Python 2 Unicode objects will likely never be supported.

2.6.2.4. tuple¶

The following operations are supported:

• tuple construction
• tuple unpacking
• comparison between tuples
• iteration and indexing over homogeneous tuples
• addition (concatenation) between tuples
• slicing tuples with a constant slice
• the index method on tuples

2.6.2.5. list¶

Creating and returning lists from JIT-compiled functions is supported, as well as all methods and operations. Lists must be strictly homogeneous: Numba will reject any list containing objects of different types, even if the types are compatible (for example, [1, 2.5] is rejected as it contains a int and a float).

For example, to create a list of arrays:

In [1]: from numba import njit

In [2]: import numpy as np

In [3]: @njit
  ...: def foo(x):
  ...:     lst = []
  ...:     for i in range(x):
  ...:         lst.append(np.arange(i))
  ...:     return lst
  ...:

In [4]: foo(4)
Out[4]: [array([], dtype=int64), array([0]), array([0, 1]), array([0, 1, 2])]

In [1]: from numba import njit

In [2]: @njit
  ...: def foo(x):
  ...:     return [[i for i in range(n)] for n in range(x)]
  ...:

In [3]: foo(3)
Out[3]: [[], [0], [0, 1]]

from numba import jit
import numpy as np

@jit(nopython=True)
def f(n):
  return np.array([ [ x * y for x in range(n) ] for y in range(n) ])

2.6.2.6. set¶

All methods and operations on sets are supported in JIT-compiled functions.

Sets must be strictly homogeneous: Numba will reject any set containing objects of different types, even if the types are compatible (for example, {1, 2.5} is rejected as it contains a int and a float).

2.6.2.7. None¶

The None value is supported for identity testing (when using an optional type).

2.6.2.8. bytes, bytearray, memoryview¶

The bytearray type and, on Python 3, the bytes type support indexing, iteration and retrieving the len().

The memoryview type supports indexing, slicing, iteration, retrieving the len(), and also the following attributes:

• contiguous
• c_contiguous
• f_contiguous
• itemsize
• nbytes
• ndim
• readonly
• shape
• strides

2.6.4.1. array¶

Limited support for the array.array type is provided through the buffer protocol. Indexing, iteration and taking the len() is supported. All type codes are supported except for "u".

2.6.4.2. cmath¶

The following functions from the cmath module are supported:

• cmath.acos()
• cmath.acosh()
• cmath.asin()
• cmath.asinh()
• cmath.atan()
• cmath.atanh()
• cmath.cos()
• cmath.cosh()
• cmath.exp()
• cmath.isfinite()
• cmath.isinf()
• cmath.isnan()
• cmath.log()
• cmath.log10()
• cmath.phase()
• cmath.polar()
• cmath.rect()
• cmath.sin()
• cmath.sinh()
• cmath.sqrt()
• cmath.tan()
• cmath.tanh()

2.6.4.3. collections¶

Named tuple classes, as returned by collections.namedtuple(), are supported in the same way regular tuples are supported. Attribute access and named parameters in the constructor are also supported.

Creating a named tuple class inside Numba code is not supported; the class must be created at the global level.

2.6.4.4. ctypes¶

Numba is able to call ctypes-declared functions with the following argument and return types:

• ctypes.c_int8
• ctypes.c_int16
• ctypes.c_int32
• ctypes.c_int64
• ctypes.c_uint8
• ctypes.c_uint16
• ctypes.c_uint32
• ctypes.c_uint64
• ctypes.c_float
• ctypes.c_double
• ctypes.c_void_p

2.6.4.5. enum¶

Both enum.Enum and enum.IntEnum subclasses are supported.

2.6.4.6. math¶

The following functions from the math module are supported:

• math.acos()
• math.acosh()
• math.asin()
• math.asinh()
• math.atan()
• math.atan2()
• math.atanh()
• math.ceil()
• math.copysign()
• math.cos()
• math.cosh()
• math.degrees()
• math.erf()
• math.erfc()
• math.exp()
• math.expm1()
• math.fabs()
• math.floor()
• math.frexp()
• math.gamma()
• math.hypot()
• math.isfinite()
• math.isinf()
• math.isnan()
• math.ldexp()
• math.lgamma()
• math.log()
• math.log10()
• math.log1p()
• math.pow()
• math.radians()
• math.sin()
• math.sinh()
• math.sqrt()
• math.tan()
• math.tanh()
• math.trunc()

2.6.4.7. operator¶

The following functions from the operator module are supported:

• operator.add()
• operator.and_()
• operator.div() (Python 2 only)
• operator.eq()
• operator.floordiv()
• operator.ge()
• operator.gt()
• operator.iadd()
• operator.iand()
• operator.idiv() (Python 2 only)
• operator.ifloordiv()
• operator.ilshift()
• operator.imatmul() (Python 3.5 and above)
• operator.imod()
• operator.imul()
• operator.invert()
• operator.ior()
• operator.ipow()
• operator.irshift()
• operator.isub()
• operator.itruediv()
• operator.ixor()
• operator.le()
• operator.lshift()
• operator.lt()
• operator.matmul() (Python 3.5 and above)
• operator.mod()
• operator.mul()
• operator.ne()
• operator.neg()
• operator.not_()
• operator.or_()
• operator.pos()
• operator.pow()
• operator.rshift()
• operator.sub()
• operator.truediv()
• operator.xor()

2.6.4.8. functools¶

The functools.reduce() function is supported but the initializer argument is required.

2.6.4.9. random¶

Numba supports top-level functions from the random module, but does not allow you to create individual Random instances. A Mersenne-Twister generator is used, with a dedicated internal state. It is initialized at startup with entropy drawn from the operating system.

• random.betavariate()
• random.expovariate()
• random.gammavariate()
• random.gauss()
• random.getrandbits(): number of bits must not be greater than 64
• random.lognormvariate()
• random.normalvariate()
• random.paretovariate()
• random.randint()
• random.random()
• random.randrange()
• random.seed(): with an integer argument only
• random.shuffle(): the sequence argument must be a one-dimension Numpy array or buffer-providing object (such as a bytearray or array.array); the second (optional) argument is not supported
• random.uniform()
• random.triangular()
• random.vonmisesvariate()
• random.weibullvariate()

2.6.5.1. cffi¶

Similarly to ctypes, Numba is able to call into cffi-declared external functions, using the following C types and any derived pointer types:

• char
• short
• int
• long
• long long
• unsigned char
• unsigned short
• unsigned int
• unsigned long
• unsigned long long
• int8_t
• uint8_t
• int16_t
• uint16_t
• int32_t
• uint32_t
• int64_t
• uint64_t
• float
• double
• ssize_t
• size_t
• void

The from_buffer() method of cffi.FFI and CompiledFFI objects is supported for passing Numpy arrays and other buffer-like objects. Only contiguous arguments are accepted. The argument to from_buffer() is converted to a raw pointer of the appropriate C type (for example a double * for a float64 array).

Additional type mappings for the conversion from a buffer to the appropriate C type may be registered with Numba. This may include struct types, though it is only permitted to call functions that accept pointers to structs - passing a struct by value is unsupported. For registering a mapping, use:

numba.cffi_support.register_type(cffi_type, numba_type)¶

Out-of-line cffi modules must be registered with Numba prior to the use of any of their functions from within Numba-compiled functions:

numba.cffi_support.register_module(mod)¶

Register the cffi out-of-line module mod with Numba.

Inline cffi modules require no registration.