import numpy as np
from ..strategy import ConnectionStrategy
[docs]class ConnectomeGolgiGranule(ConnectionStrategy):
"""
Legacy implementation for the connections between Golgi cells and granule cells.
"""
casts = {"detailed": bool}
defaults = {"detailed": False}
[docs] def validate(self):
if self.detailed:
morphologies = self.from_cell_types[0].list_all_morphologies()
if not morphologies:
raise ConfigurationError(
"Can't create detailed golgi to granule connections without any morphologies for the golgi cell."
)
elif len(morphologies) > 1:
raise NotImplementedError(
"Detailed golgi to granule connections can only be made for a single golgi morphology."
+ " (Requires the selection of morphologies to be moved from the connection module to the placement module)"
)
mr = self.scaffold.morphology_repository
morphology = mr.get_morphology(morphologies[0])
axonic_compartments = morphology.get_compartments(["axon"])
self.axon = np.array([c.id for c in axonic_compartments])
self.morphology = morphology
def connect(self):
# Gather information for the legacy code block below.
glom_grc = self.scaffold.cell_connections_by_tag["glomerulus_to_granule"]
glom_ids = self.scaffold.get_placement_set("glomerulus").identifiers
goc_glom = self.scaffold.cell_connections_by_tag["golgi_to_glomerulus"]
golgi_type = self.from_cell_types[0]
golgis = self.scaffold.cells_by_type[golgi_type.name]
if self.detailed:
# Fetch the connection compartments of glomerulus to granule connections.
try:
glom_grc_compartments = self.scaffold.connection_compartments[
"glomerulus_to_granule"
]
except KeyError:
raise RuntimeError(
"Missing the glomerulus to granule connection compartments."
)
# Index the glom to granule dense matrix to an adjacency list
glom_target_map = {gid: [] for gid in glom_ids}
for glom, grc in glom_grc:
glom_target_map[glom].append(grc)
if self.detailed:
# Lookup the grc compartments associated with each glom-grc conn
glom_comp_map = {gid: [] for gid in glom_ids}
for i, glom in enumerate(glom_grc[:, 0]):
glom_comp_map[glom].append(glom_grc_compartments[i, 1])
# Lookup the grc targets of each golgi-glom in the order they appear.
grc_via_glom = np.vectorize(glom_target_map.get, otypes=[np.ndarray])(
goc_glom[:, 1]
)
if self.detailed:
comp_via_glom = np.vectorize(glom_comp_map.get, otypes=[np.ndarray])(
goc_glom[:, 1]
)
# Sum the targets to preallocate total size
malloc = sum(map(len, grc_via_glom))
connections = np.empty((malloc, 2))
ptr = 0
# Iterate over the golgis and their grc targets (as lookup up in order
# via the gloms) and fill them into the connections array
for goc, grcs in zip(goc_glom[:, 0], grc_via_glom):
s = len(grcs)
if not s:
# Skip any weird slicing of empty target sets
continue
connections[ptr : (ptr + s), 0] = goc
connections[ptr : (ptr + s), 1] = grcs
ptr += s
if self.detailed:
compartments = np.empty((malloc, 2))
# Assign random axonal segments
compartments[:, 0] = np.random.choice(self.axon, malloc)
ptr = 0
# Assign the glom associated dendrites
for comps in comp_via_glom:
s = len(comps)
if not s:
# Skip any weird slicing of empty target sets
continue
compartments[ptr : (ptr + s), 1] = comps
ptr += s
# Make a bad map ignoring all but the first (only) morphologies
# of each type. Sorry future me <3
morpho_map = [
self.from_cell_types[0].list_all_morphologies()[0],
self.to_cell_types[0].list_all_morphologies()[0],
]
morphologies = np.zeros((len(compartments), 2))
morphologies[:, 1] = 1
self.scaffold.connect_cells(
self,
connections,
morphologies=morphologies,
compartments=compartments,
morpho_map=morpho_map,
)
else:
self.scaffold.connect_cells(self, connections)
