neurokernel.pm_gpu.GPUPortMapper

class neurokernel.pm_gpu.GPUPortMapper(selector, data=None, portmap=None, make_copy=True)

Maps a PyCUDA GPUArray to/from path-like port identifiers.

__init__(selector, data=None, portmap=None, make_copy=True)

x.__init__(…) initializes x; see help(type(x)) for signature

Methods

__init__(selector[, data, portmap, make_copy])	x.__init__(…) initializes x; see help(type(x)) for signature
copy()	Return copy of this port mapper.
equals(other)	Check whether this mapper is equivalent to another mapper.
from_index(idx, data[, portmap])	Create port mapper from a Pandas index and a sequence of integer indices.
from_pm(pm)	Create a new port mapper instance given an existing instance.
get(selector)	Retrieve mapped data specified by given selector.
get_by_inds(inds)	Retrieve mapped data specified by integer index.
get_inds_nonzero()	Select indices of ports with nonzero data.
get_map(selector)	Retrieve integer indices associated with selector.
get_ports(f)	Select ports using a data selection function.
get_ports_as_inds(f)	Select integer indices corresponding to ports in map.
get_ports_nonzero()	Select ports with nonzero data.
inds_to_ports(inds)	Convert list of integer indices to port identifiers.
ports_to_inds(selector)	Convert port selector to list of integer indices.
set(selector, data)	Set mapped data specified by given selector.
set_by_inds(inds, data)	Set mapped data by integer indices.
set_by_inds_array(inds, data)	Set mapped data with array by integer indices.
set_by_inds_scalar(inds, data)	Set mapped data with scalar by integer indices.
set_map(selector, portmap)	Set mapped integer index associated with selector.

Attributes

data	Data associated with ports.
data_ctype	C type corresponding to type of data array.
dtype	Port mapper data type.
index	Port mapper index.