mpi4py.MPI.Datatype

class mpi4py.MPI.Datatype(datatype=None)

Bases: object

Datatype object

Parameters: datatype (Optional[Datatype]) –
Return type: Datatype

static __new__(cls, datatype=None)

Parameters: datatype (Optional[Datatype]) –
Return type: Datatype

Methods Summary

`Commit`()	Commit the datatype
`Create_contiguous`(count)	Create a contiguous datatype
`Create_darray`(size, rank, gsizes, distribs, ...)	Create a datatype representing an HPF-like distributed array on Cartesian process grids
`Create_f90_complex`(p, r)	Return a bounded complex datatype
`Create_f90_integer`(r)	Return a bounded integer datatype
`Create_f90_real`(p, r)	Return a bounded real datatype
`Create_hindexed`(blocklengths, displacements)	Create an indexed datatype with displacements in bytes
`Create_hindexed_block`(blocklength, displacements)	Create an indexed datatype with constant-sized blocks and displacements in bytes
`Create_hvector`(count, blocklength, stride)	Create a vector (strided) datatype
`Create_indexed`(blocklengths, displacements)	Create an indexed datatype
`Create_indexed_block`(blocklength, displacements)	Create an indexed datatype with constant-sized blocks
`Create_keyval`([copy_fn, delete_fn, nopython])	Create a new attribute key for datatypes
`Create_resized`(lb, extent)	Create a datatype with a new lower bound and extent
`Create_struct`(blocklengths, displacements, ...)	Create an datatype from a general set of block sizes, displacements and datatypes
`Create_subarray`(sizes, subsizes, starts[, order])	Create a datatype for a subarray of a regular, multidimensional array
`Create_vector`(count, blocklength, stride)	Create a vector (strided) datatype
`Delete_attr`(keyval)	Delete attribute value associated with a key
`Dup`()	Duplicate a datatype
`Free`()	Free the datatype
`Free_keyval`(keyval)	Free an attribute key for datatypes
`Get_attr`(keyval)	Retrieve attribute value by key
`Get_contents`()	Retrieve the actual arguments used in the call that created a datatype
`Get_envelope`()	Return information on the number and type of input arguments used in the call that created a datatype
`Get_extent`()	Return lower bound and extent of datatype
`Get_name`()	Get the print name for this datatype
`Get_size`()	Return the number of bytes occupied by entries in the datatype
`Get_true_extent`()	Return the true lower bound and extent of a datatype
`Match_size`(typeclass, size)	Find a datatype matching a specified size in bytes
`Pack`(inbuf, outbuf, position, comm)	Pack into contiguous memory according to datatype.
`Pack_external`(datarep, inbuf, outbuf, position)	Pack into contiguous memory according to datatype, using a portable data representation (external32).
`Pack_external_size`(datarep, count)	Return the upper bound on the amount of space (in bytes) needed to pack a message according to datatype, using a portable data representation (external32).
`Pack_size`(count, comm)	Return the upper bound on the amount of space (in bytes) needed to pack a message according to datatype.
`Set_attr`(keyval, attrval)	Store attribute value associated with a key
`Set_name`(name)	Set the print name for this datatype
`Unpack`(inbuf, position, outbuf, comm)	Unpack from contiguous memory according to datatype.
`Unpack_external`(datarep, inbuf, position, outbuf)	Unpack from contiguous memory according to datatype, using a portable data representation (external32).
`decode`()	Convenience method for decoding a datatype
`f2py`(arg)
`py2f`()

Attributes Summary

`combiner`	datatype combiner
`contents`	datatype contents
`envelope`	datatype envelope
`extent`
`is_named`	is a named datatype
`is_predefined`	is a predefined datatype
`lb`	lower bound
`name`	datatype name
`size`
`true_extent`	true extent
`true_lb`	true lower bound
`true_ub`	true upper bound
`ub`	upper bound

Methods Documentation

Commit()

Commit the datatype

Return type: Datatype

Create_contiguous(count)

Create a contiguous datatype

Parameters: count (int) –
Return type: Datatype

Create_darray(size, rank, gsizes, distribs, dargs, psizes, order=ORDER_C)

Create a datatype representing an HPF-like distributed array on Cartesian process grids

Parameters

size (int) –
rank (int) –
gsizes (Sequence[int]) –
distribs (Sequence[int]) –
dargs (Sequence[int]) –
psizes (Sequence[int]) –
order (int) –

Return type

Datatype

classmethod Create_f90_complex(p, r)

Return a bounded complex datatype

Parameters

p (int) –
r (int) –

Return type

Datatype

classmethod Create_f90_integer(r)

Return a bounded integer datatype

Parameters: r (int) –
Return type: Datatype

classmethod Create_f90_real(p, r)

Return a bounded real datatype

Parameters

p (int) –
r (int) –

Return type

Datatype

Create_hindexed(blocklengths, displacements)

Create an indexed datatype with displacements in bytes

Parameters

blocklengths (Sequence[int]) –
displacements (Sequence[int]) –

Return type

Datatype

Create_hindexed_block(blocklength, displacements)

Create an indexed datatype with constant-sized blocks and displacements in bytes

Parameters

blocklength (int) –
displacements (Sequence[int]) –

Return type

Datatype

Create_hvector(count, blocklength, stride)

Create a vector (strided) datatype

Parameters

count (int) –
blocklength (int) –
stride (int) –

Return type

Datatype

Create_indexed(blocklengths, displacements)

Create an indexed datatype

Parameters

blocklengths (Sequence[int]) –
displacements (Sequence[int]) –

Return type

Datatype

Create_indexed_block(blocklength, displacements)

Create an indexed datatype with constant-sized blocks

Parameters

blocklength (int) –
displacements (Sequence[int]) –

Return type

Datatype

classmethod Create_keyval(copy_fn=None, delete_fn=None, nopython=False)

Create a new attribute key for datatypes

Parameters

copy_fn (Optional[Callable[[Datatype, int, Any], Any]]) –
delete_fn (Optional[Callable[[Datatype, int, Any], None]]) –
nopython (bool) –

Return type

int

Create_resized(lb, extent)

Create a datatype with a new lower bound and extent

Parameters

lb (int) –
extent (int) –

Return type

Datatype

classmethod Create_struct(blocklengths, displacements, datatypes)

Create an datatype from a general set of block sizes, displacements and datatypes

Parameters

blocklengths (Sequence[int]) –
displacements (Sequence[int]) –
datatypes (Sequence[Datatype]) –

Return type

Datatype

Create_subarray(sizes, subsizes, starts, order=ORDER_C)

Create a datatype for a subarray of a regular, multidimensional array

Parameters

sizes (Sequence[int]) –
subsizes (Sequence[int]) –
starts (Sequence[int]) –
order (int) –

Return type

Datatype

Create_vector(count, blocklength, stride)

Create a vector (strided) datatype

Parameters

count (int) –
blocklength (int) –
stride (int) –

Return type

Datatype

Delete_attr(keyval)

Delete attribute value associated with a key

Parameters: keyval (int) –
Return type: None

Dup()

Duplicate a datatype

Return type: Datatype

Free()

Free the datatype

Return type: None

classmethod Free_keyval(keyval)

Free an attribute key for datatypes

Parameters: keyval (int) –
Return type: int

Get_attr(keyval)

Retrieve attribute value by key

Parameters: keyval (int) –
Return type: Optional[Union[int, Any]]

Get_contents()

Retrieve the actual arguments used in the call that created a datatype

Return type: Tuple[List[int], List[int], List[Datatype]]

Get_envelope()

Return information on the number and type of input arguments used in the call that created a datatype

Return type: Tuple[int, int, int, int]

Get_extent()

Return lower bound and extent of datatype

Return type: Tuple[int, int]

Get_name()

Get the print name for this datatype

Return type: str

Get_size()

Return the number of bytes occupied by entries in the datatype

Return type: int

Get_true_extent()

Return the true lower bound and extent of a datatype

Return type: Tuple[int, int]

classmethod Match_size(typeclass, size)

Find a datatype matching a specified size in bytes

Parameters

typeclass (int) –
size (int) –

Return type

Datatype

Pack(inbuf, outbuf, position, comm)

Pack into contiguous memory according to datatype.

Parameters

inbuf (BufSpec) –
outbuf (BufSpec) –
position (int) –
comm (Comm) –

Return type

int

Pack_external(datarep, inbuf, outbuf, position)

Pack into contiguous memory according to datatype, using a portable data representation (external32).

Parameters

datarep (str) –
inbuf (BufSpec) –
outbuf (BufSpec) –
position (int) –

Return type

int

Pack_external_size(datarep, count)

Return the upper bound on the amount of space (in bytes) needed to pack a message according to datatype, using a portable data representation (external32).

Parameters

datarep (str) –
count (int) –

Return type

int

Pack_size(count, comm)

Return the upper bound on the amount of space (in bytes) needed to pack a message according to datatype.

Parameters

count (int) –
comm (Comm) –

Return type

int

Set_attr(keyval, attrval)

Store attribute value associated with a key

Parameters

keyval (int) –
attrval (Any) –

Return type

None

Set_name(name)

Set the print name for this datatype

Parameters: name (str) –
Return type: None

Unpack(inbuf, position, outbuf, comm)

Unpack from contiguous memory according to datatype.

Parameters

inbuf (BufSpec) –
position (int) –
outbuf (BufSpec) –
comm (Comm) –

Return type

int

Unpack_external(datarep, inbuf, position, outbuf)

Unpack from contiguous memory according to datatype, using a portable data representation (external32).

Parameters

datarep (str) –
inbuf (BufSpec) –
position (int) –
outbuf (BufSpec) –

Return type

int

decode()

Convenience method for decoding a datatype

Return type: Tuple[Datatype, str, Dict[str, Any]]

classmethod f2py(arg)

Parameters: arg (int) –
Return type: Datatype

py2f()

Return type: int

Attributes Documentation

combiner: datatype combiner

contents: datatype contents

envelope: datatype envelope

extent

is_named: is a named datatype

is_predefined: is a predefined datatype

lb: lower bound

name: datatype name

size

true_extent: true extent

true_lb: true lower bound

true_ub: true upper bound

ub: upper bound