PythonCall API Reference

Py(x)

Convert x to a Python object, of type Py.

Conversion happens according to these rules.

Such an object supports attribute access (obj.attr), indexing (obj[idx]), calling (obj(arg1, arg2)), iteration (for x in obj), arithmetic (obj + obj2) and comparison (obj > obj2), among other things. These operations convert all their arguments to Py and return Py.

pybuiltins

An object whose fields are the Python builtins, of type Py.

For example pybuiltins.None, pybuiltins.int, pybuiltins.ValueError.

pybool(x)

Convert x to a Python bool.

pycollist(x::AbstractArray)

Create a nested Python list-of-lists from the elements of x. For matrices, this is a list of columns.

pybytes(x)

Convert x to a Python bytes.

pycomplex(x=0.0)
pycomplex(re, im)

Convert x to a Python complex, or create one from given real and imaginary parts.

pydict(x)
pydict(; x...)

Convert x to a Python dict. In the second form, the keys are strings.

If x is a Python object, this is equivalent to dict(x) in Python. Otherwise x must iterate over key-value pairs.

pyfloat(x=0.0)

Convert x to a Python float.

pyfrozenset(x=())

Convert x to a Python frozenset.

If x is a Python object, this is equivalent to frozenset(x) in Python. Otherwise x must be iterable.

pyint(x=0)

Convert x to a Python int.

pylist(x=())

Convert x to a Python list.

If x is a Python object, this is equivalent to list(x) in Python. Otherwise x must be iterable.

pyrange([[start], [stop]], [step])

Construct a Python range. Unspecified arguments default to None.

pyrowlist(x::AbstractArray)

Create a nested Python list-of-lists from the elements of x. For matrices, this is a list of rows.

pyset(x=())

Convert x to a Python set.

If x is a Python object, this is equivalent to set(x) in Python. Otherwise x must be iterable.

pyslice([start], stop, [step])

Construct a Python slice. Unspecified arguments default to None.

pystr(x)

Convert x to a Python str.

pytuple(x=())

Convert x to a Python tuple.

If x is a Python object, this is equivalent to tuple(x) in Python. Otherwise x must be iterable.

pyall(x)

Equivalent to all(x) in Python.

pyany(x)

Equivalent to any(x) in Python.

pyascii(x)

Equivalent to ascii(x) in Python.

pycall(f, args...; kwargs...)

Call the Python object f with the given arguments.

pycallable(x)

Equivalent to callable(x) in Python.

pycompile(...)

Equivalent to compile(...) in Python.

pycontains(x, v)

Equivalent to v in x in Python.

pydelattr(x, k)

Equivalent to delattr(x, k) or del x.k in Python.

pydelitem(x, k)

Equivalent to delitem(x, k) or del x[k] in Python.

pydir(x)

Equivalent to dir(x) in Python.

pyeval([T=Py], code, globals, locals=nothing)

Evaluate the given Python code, returning the result as a T.

If globals is a Module, then a persistent dict unique to that module is used.

By default the code runs in global scope (i.e. locals===globals). To use a temporary local scope, set locals to (), or to a NamedTuple of variables to include in the scope.

See also @pyeval.

Examples

The following computes 1.1+2.2 in the Main module as a Float64:

pyeval(Float64, "x+y", Main, (x=1.1, y=2.2))  # returns 3.3

@pyeval [inputs =>] code [=> T]

Evaluate the given code in a new local scope and return the answer as a T.

The global scope is persistent and unique to the current module.

The code must be a literal string or command.

The inputs is a tuple of inputs of the form v=expr to be included in the local scope. Only v is required, expr defaults to v.

Examples

The following computes 1.1+2.2 and returns a Float64:

@pyeval (x=1.1, y=2.2) => `x+y` => Float64  # returns 3.3

pyexec([T=Nothing], code, globals, locals=nothing)

Execute the given Python code.

If globals is a Module, then a persistent dict unique to that module is used.

By default the code runs in global scope (i.e. locals===globals). To use a temporary local scope, set locals to (), or to a NamedTuple of variables to include in the scope.

If T==Nothing then returns nothing. Otherwise T must be a concrete NamedTuple type and the corresponding items from locals are extracted and returned.

See also @pyexec.

Examples

The following computes 1.1+2.2 in the Main module as a Float64:

pyexec(@NamedTuple{ans::Float64}, "ans=x+y", Main, (x=1.1, y=2.2))  # returns (ans = 3.3,)

Marking variables as global saves them into the module scope, so that they are available in subsequent invocations:

pyexec("global x; x=12", Main)
pyeval(Int, "x", Main)  # returns 12

@pyexec [inputs =>] code [=> outputs]

Execute the given code in a new local scope.

The global scope is persistent and unique to the current module.

The code must be a literal string or command.

The inputs is a tuple of inputs of the form v=expr to be included in the local scope. Only v is required, expr defaults to v.

The outputs is a tuple of outputs of the form x::T=v, meaning that v is extracted from locals, converted to T and assigned to x. Only x is required: T defaults to Py and v defaults to x.

Examples

The following computes 1.1+2.2 and assigns its value to ans as a Float64:

@pyexec (x=1.1, y=2.2) => `ans=x+y` => ans::Float64  # returns 3.3

Marking variables as global saves them into the module scope, so that they are available in subsequent invocations:

@pyexec `global x; x=12`
@pyeval `x` => Int  # returns 12

pygetattr(x, k, [d])

Equivalent to getattr(x, k) or x.k in Python.

If d is specified, it is returned if the attribute does not exist.

pygetitem(x, k, [d])

Equivalent x[k] in Python.

If d is specified, it is returned if the item does not exist (i.e. if x[k] raises a KeyError or IndexError).

pyhasattr(x, k)

Equivalent to hasattr(x, k) in Python.

Tests if getattr(x, k) raises an AttributeError.

pyhasitem(x, k)

Test if pygetitem(x, k) raises a KeyError or AttributeError.

pyhash(x)

Equivalent to hash(x) in Python, converted to an Integer.

pyhelp([x])

Equivalent to help(x) in Python.

pyimport(m)
pyimport(m => k)
pyimport(m => (k1, k2, ...))
pyimport(m1, m2, ...)

Import a module m, or an attribute k, or a tuple of attributes.

If several arguments are given, return the results of importing each one in a tuple.

pyin(v, x)

Equivalent to v in x in Python.

pyis(x, y)

True if x and y are the same Python object. Equivalent to x is y in Python.

pyisinstance(x, t)

Test if x is of type t. Equivalent to isinstance(x, t) in Python.

pyissubclass(s, t)

Test if s is a subclass of t. Equivalent to issubclass(s, t) in Python.

pyiter(x)

Equivalent to iter(x) in Python.

pylen(x)

The length of x. Equivalent to len(x) in Python, converted to an Integer.

pynext(x)

Equivalent to next(x) in Python.

pyprint(...)

Equivalent to print(...) in Python.

pyrepr(x)

Equivalent to repr(x) in Python.

pysetattr(x, k, v)

Equivalent to setattr(x, k, v) or x.k = v in Python.

pysetitem(x, k, v)

Equivalent to setitem(x, k, v) or x[k] = v in Python.

pytype(x)

The Python type of x.

pywith(f, o, d=nothing)

Equivalent to with o as x: f(x) in Python, where x is a Py.

On success, the value of f(x) is returned.

If an exception occurs but is suppressed then d is returned.

pyconvert(T, x, [d])

Convert the Python object x to a T.

If d is specified, it is returned on failure instead of throwing an error.

@pyconvert(T, x, [onfail])

Convert the Python object x to a T.

On failure, evaluates to onfail, which defaults to return pyconvert_unconverted() (mainly useful for writing conversion rules).

pyjl([t=pyjltype(x)], x)

Create a Python object wrapping the Julia object x.

If x is mutable, then mutating the returned object also mutates x, and vice versa.

Its Python type is normally inferred from the type of x, but can be specified with t.

For example if x is an AbstractVector then the object will have type juliacall.VectorValue. This object will satisfy the Python sequence interface, so for example uses 0-up indexing.

To define a custom conversion for your type T, overload pyjltype(::T).

pyjlraw(v)

Create a Python object wrapping the Julia object x.

It has type juliacall.RawValue. This has a much more rigid "Julian" interface than pyjl(v). For example, accessing attributes or calling this object will always return a RawValue.

pyisjl(x)

Test whether x is a wrapped Julia value, namely an instance of juliacall.ValueBase.

pyjlvalue(x)

Extract the value from the wrapped Julia value x.

pybinaryio(io::IO)

Wrap io as a Python binary IO object.

This is the default behaviour of Py(io).

pytextio(io::IO)

Wrap io as a Python text IO object.

pyneg(x)

Equivalent to -x in Python.

pypos(x)

Equivalent to +x in Python.

pyabs(x)

Equivalent to abs(x) in Python.

pyinv(x)

Equivalent to ~x in Python.

pyindex(x)

Convert x losslessly to an int.

pyadd(x, y)

Equivalent to x + y in Python.

pysub(x, y)

Equivalent to x - y in Python.

pymul(x, y)

Equivalent to x * y in Python.

pymatmul(x, y)

Equivalent to x @ y in Python.

pypow(x, y, z=None)

Equivalent to x ** y or pow(x, y, z) in Python.

pyfloordiv(x, y)

Equivalent to x // y in Python.

pytruediv(x, y)

Equivalent to x / y in Python.

pymod(x, y)

Equivalent to x % y in Python.

pydivmod(x, y)

Equivalent to divmod(x, y) in Python.

pylshift(x, y)

Equivalent to x << y in Python.

pyrshift(x, y)

Equivalent to x >> y in Python.

pyand(x, y)

Equivalent to x & y in Python.

pyxor(x, y)

Equivalent to x ^ y in Python.

pyor(x, y)

Equivalent to x | y in Python.

pyiadd(x, y)

In-place add. x = pyiadd(x, y) is equivalent to x += y in Python.

pyisub(x, y)

In-place subtract. x = pyisub(x, y) is equivalent to x -= y in Python.

pyimul(x, y)

In-place multiply. x = pyimul(x, y) is equivalent to x *= y in Python.

pyimatmul(x, y)

In-place matrix multiply. x = pyimatmul(x, y) is equivalent to x @= y in Python.

pyipow(x, y, z=None)

In-place power. x = pyipow(x, y) is equivalent to x **= y in Python.

pyifloordiv(x, y)

In-place floor divide. x = pyifloordiv(x, y) is equivalent to x //= y in Python.

pyitruediv(x, y)

In-place true division. x = pyitruediv(x, y) is equivalent to x /= y in Python.

pyimod(x, y)

In-place subtraction. x = pyimod(x, y) is equivalent to x %= y in Python.

pyilshift(x, y)

In-place left shift. x = pyilshift(x, y) is equivalent to x <<= y in Python.

pyirshift(x, y)

In-place right shift. x = pyirshift(x, y) is equivalent to x >>= y in Python.

pyiand(x, y)

In-place and. x = pyiand(x, y) is equivalent to x &= y in Python.

pyixor(x, y)

In-place xor. x = pyixor(x, y) is equivalent to x ^= y in Python.

pyior(x, y)

In-place or. x = pyior(x, y) is equivalent to x |= y in Python.

pytruth(x)

The truthyness of x. Equivalent to bool(x) in Python, converted to a Bool.

pynot(x)

The falsyness of x. Equivalent to not x in Python, converted to a Bool.

pyeq(x, y)
pyeq(Bool, x, y)

Equivalent to x == y in Python. The second form converts to Bool.

pyne(x, y)
pyne(Bool, x, y)

Equivalent to x != y in Python. The second form converts to Bool.

pyle(x, y)
pyle(Bool, x, y)

Equivalent to x <= y in Python. The second form converts to Bool.

pylt(x, y)
pylt(Bool, x, y)

Equivalent to x < y in Python. The second form converts to Bool.

pyge(x, y)
pyge(Bool, x, y)

Equivalent to x >= y in Python. The second form converts to Bool.

pygt(x, y)
pygt(Bool, x, y)

Equivalent to x > y in Python. The second form converts to Bool.

pytype(name, bases, dict)

Create a new type. Equivalent to type(name, bases, dict) in Python.

If bases is not a Python object, it is converted to one using pytuple.

The dict may either by a Python object or a Julia iterable. In the latter case, each item may either be a name => value pair or a Python object with a __name__ attribute.

In order to use a Julia Function as an instance method, it must be wrapped into a Python function with pyfunc. Similarly, see also pyclassmethod, pystaticmethod or pyproperty. In all these cases, the arguments passed to the function always have type Py. See the example below.

Example

Foo = pytype("Foo", (), [
    "__module__" => "__main__",

    pyfunc(
        name = "__init__",
        doc = """
        Specify x and y to store in the Foo.

        If omitted, y defaults to None.
        """,
        function (self, x, y = nothing)
            self.x = x
            self.y = y
            return
        end,
    ),

    pyfunc(
        name = "__repr__",
        self -> "Foo($(self.x), $(self.y))",
    ),

    pyclassmethod(
        name = "frompair",
        doc = "Construct a Foo from a tuple of length two.",
        (cls, xy) -> cls(xy...),
    ),

    pystaticmethod(
        name = "hello",
        doc = "Prints a friendly greeting.",
        (name) -> println("Hello, $name"),
    ),

    "xy" => pyproperty(
        doc = "A tuple of x and y.",
        get = (self) -> (self.x, self.y),
        set = function (self, xy)
            (x, y) = xy
            self.x = x
            self.y = y
            nothing
        end,
    ),
])

pyfunc(f; [name], [qualname], [doc], [signature])

Wrap the callable f as an ordinary Python function.

The name, qualname, docstring or signature can optionally be set with name, qualname, doc or signature.

Unlike Py(f) (or pyjl(f)), the arguments passed to f are always of type Py, i.e. they are never converted.

pyclassmethod(f; ...)

Convert callable f to a Python class method.

If f is not a Python object (e.g. if f is a Function) then it is converted to one with pyfunc. In particular this means the arguments passed to f are always of type Py. Keyword arguments are passed to pyfunc.

pystaticmethod(f; ...)

Convert callable f to a Python static method.

If f is not a Python object (e.g. if f is a Function) then it is converted to one with pyfunc. In particular this means the arguments passed to f are always of type Py. Any keyword arguments are passed to pyfunc.

pyproperty(; get=nothing, set=nothing, del=nothing, doc=nothing)
pyproperty(get)

Create a Python property with the given getter, setter and deleter.

If get, set or del is not a Python object (e.g. if it is a Function) then it is converted to one with pyfunc. In particular this means the arguments passed to it are always of type Py.

PyList{T=Py}([x])

Wraps the Python list x (or anything satisfying the sequence interface) as an AbstractVector{T}.

If x is not a Python object, it is converted to one using pylist.

PySet{T=Py}([x])

Wraps the Python set x (or anything satisfying the set interface) as an AbstractSet{T}.

If x is not a Python object, it is converted to one using pyset.

PyDict{K=Py,V=Py}([x])

Wraps the Python dict x (or anything satisfying the mapping interface) as an AbstractDict{K,V}.

If x is not a Python object, it is converted to one using pydict.

PyIterable{T=Py}(x)

This object iterates over iterable Python object x, yielding values of type T.

PyArray{T,N,M,L,R}(x; copy=true, array=true, buffer=true)

Wrap the Python array x as a Julia AbstractArray{T,N}.

The input x can be bytes, bytearray, array.array, numpy.ndarray or anything satisfying the buffer protocol (if buffer=true) or the numpy array interface (if array=true).

If copy=false then the resulting array is guaranteed to directly wrap the data in x. If copy=true then a copy is taken if necessary to produce an array.

The type parameters are all optional, and are:

• T: The element type.
• N: The number of dimensions.
• M: True if the array is mutable.
• L: True if the array supports fast linear indexing.
• R: The element type of the underlying buffer. Often equal to T.

PyIO(x; own=false, text=missing, line_buffering=false, buflen=4096)

Wrap the Python IO stream x as a Julia IO stream.

When this goes out of scope and is finalized, it is automatically flushed. If own=true then it is also closed.

If text=false then x must be a binary stream and arbitrary binary I/O is possible. If text=true then x must be a text stream and only UTF-8 must be written (i.e. use print not write). If text is not specified then it is chosen automatically. If x is a text stream and you really need a binary stream, then often PyIO(x.buffer) will work.

If line_buffering=true then output is flushed at each line.

For efficiency, reads and writes are buffered before being sent to x. The size of the buffers is buflen. The buffers are cleared using flush.

PyTable(x)

Wrap x as a Tables.jl-compatible table.

PyTable is an abstract type. See PyPandasDataFrame for a concrete example.

PyPandasDataFrame(x; [indexname::Union{Nothing,Symbol}], [columnnames::Function], [columntypes::Function])

Wraps the pandas DataFrame x as a Tables.jl-compatible table.

• indexname: The name of the column including the index. The default is nothing, meaning to exclude the index.
• columnnames: A function mapping the Python column name (a Py) to the Julia one (a Symbol). The default is x -> Symbol(x).
• columntypes: A function taking the column name (a Symbol) and returning either the desired element type of the column, or nothing to indicate automatic inference.

PyObjectArray(undef, dims...)
PyObjectArray(array)

An array of Pys which supports the Python buffer protocol.

Internally, the objects are stored as an array of pointers.

PyException(x)

Wraps the Python exception x as a Julia Exception.

@py expr

Evaluate the given expression using Pythonic semantics.

For example:

• f(x, y) is translated to pycall(f, x, y)
• x + y is translated to pyadd(x, y)
• x === y is translated to pyis(x, y) (x is y in Python)
• x.foo is translated to pygetattr(x, "foo")
• import x: f as g is translated to g = pyimport("x" => "f") (from x import f as g in Python)

Compound statements such as begin, if, while and for are supported.

See the online documentation for more details.

@pyconst ex

Equivalent to Py(ex) but always returns the exact same Julia object.

That is, if foo() = @pyconst ex then foo() === foo().

The expression ex is evaluated the first time the code is run.

If ex is a string literal, the string is interned.

Do not use this macro at the top level of a module. Instead, use pynew() and pycopy!().

lock(f)

Lock the GIL, compute f(), unlock the GIL, then return the result of f().

Use this to run Python code from threads that do not currently hold the GIL, such as new threads. Since the main Julia thread holds the GIL by default, you will need to unlock the GIL before using this function.

See @lock for the macro form.

@lock expr

Lock the GIL, compute expr, unlock the GIL, then return the result of expr.

Use this to run Python code from threads that do not currently hold the GIL, such as new threads. Since the main Julia thread holds the GIL by default, you will need to @unlock the GIL before using this function.

The macro equivalent of lock.

unlock(f)

Unlock the GIL, compute f(), re-lock the GIL, then return the result of f().

Use this to run non-Python code with the GIL unlocked, so allowing another thread to run Python code. That other thread can be a Julia thread, which must lock the GIL using lock.

See @unlock for the macro form.

@unlock expr

Unlock the GIL, compute expr, re-lock the GIL, then return the result of expr.

Use this to run non-Python code with the GIL unlocked, so allowing another thread to run Python code. That other thread can be a Julia thread, which must lock the GIL using @lock.

The macro equivalent of unlock.

PythonCall.GC.gc()

Free any Python objects waiting to be freed.

These are objects that were finalized from a thread that was not holding the Python GIL at the time.

Like most PythonCall functions, this must only be called from the main thread (i.e. the thread currently holding the Python GIL.)

python_version()

The version of Python, or missing if not known.

python_executable_path()

Path to the Python interpreter, or missing if not known.

python_library_path()

Path to libpython, or missing if not known.

python_library_handle()

Handle to the open libpython, or C_NULL if not known.

pynew([ptr])

A new Py representing the Python object at ptr (NULL by default).

If ptr is given and non-NULL, this function steals a reference to the Python object it points at, i.e. the new Py object owns a reference.

Note that NULL Python objects are not safe in the sense that most API functions will probably crash your Julia session if you pass a NULL argument.

pyisnull(x)

True if the Python object x is NULL.

pycopy!(dst::Py, src)

Copy the Python object src into dst, so that they both represent the same object.

This function exists to support module-level constant Python objects. It is illegal to call most PythonCall API functions at the top level of a module (i.e. before __init__() has run) so you cannot do const x = pything() at the top level. Instead do const x = pynew() at the top level then pycopy!(x, pything()) inside __init__().

Assumes dst is NULL, otherwise a memory leak will occur.

getptr(x)

Get the underlying pointer from the Python object x.

pydel!(x::Py)

Delete the Python object x.

DANGER! Use this function ONLY IF the Julia object x could have been garbage-collected anyway, i.e. was about to become unreachable. This means you MUST KNOW that no other part of the program has the Julia object x.

This decrements the reference count, sets the pointer to NULL and appends x to a cache of unused objects (PYNULL_CACHE).

This is an optimization to avoid excessive allocation and deallocation in Julia, which can be a significant source of slow-down in code which uses a lot of Python objects. It allows pynew() to pop an item from PYNULL_CACHE instead of allocating one, and avoids calling the relatively slow finalizer on x.

unsafe_pynext(x)

Return the next item in the iterator x. When there are no more items, return NULL.