explicit_edges

class brainpy.state.network.explicit_edges(pre_idx, post_idx)

Return a rule wiring exactly the given (pre_idx[i], post_idx[i]) edges.

A connection rule built from a precomputed edge list, for topologies that are easier to enumerate directly than to express through a sampling rule (e.g. the structured inhibitory graph of a constraint-satisfaction network). The edges are handed to the same projection paths used by the sampling rules, so a single call realizes the whole topology as one projection – avoiding the per-edge / per-group projection explosion (and the per-cont() retrace) of many small connect() calls.

Parameters:

• pre_idx (array_like of int) – Equal-length 1-D arrays of segment-local source/target neuron indices, in the coordinate frame of the populations passed to connect(). Edge i connects pre_idx[i] -> post_idx[i]. Order is preserved and duplicates are kept – the caller controls multapses.

• post_idx (array_like of int) – Equal-length 1-D arrays of segment-local source/target neuron indices, in the coordinate frame of the populations passed to connect(). Edge i connects pre_idx[i] -> post_idx[i]. Order is preserved and duplicates are kept – the caller controls multapses.

Returns:

A connection rule returning the precomputed edge set verbatim (the sampling key / autapse / multapse flags are ignored).

Return type:

_ExplicitEdges

Raises:

ValueError – If pre_idx / post_idx are not 1-D, differ in length, or are not integer-typed.

See also

brainpy.state.all_to_all, brainpy.state.Simulator.connect

Examples

>>> import numpy as np
>>> import brainunit as u
>>> from brainpy import state as bp
>>> sim = bp.Simulator(dt=0.1 * u.ms)
>>> pop = sim.create(bp.iaf_psc_exp, 5)
>>> # wire 0->1, 0->4, 2->3 as one sparse projection
>>> pre = np.array([0, 0, 2])
>>> post = np.array([1, 4, 3])
>>> _ = sim.connect(pop, pop, rule=bp.explicit_edges(pre, post),
...                 weight=-0.2 * u.pA, comm='sparse')
>>> len(sim.get_connections(source=pop, target=pop))
3