/usr/share/pyshared/pyNN/connectors.py is in python-pynn 0.7.4-1.
This file is owned by root:root, with mode 0o644.
The actual contents of the file can be viewed below.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 | """
Defines a common implementation of the built-in PyNN Connector classes.
Simulator modules may use these directly, or may implement their own versions
for improved performance.
:copyright: Copyright 2006-2011 by the PyNN team, see AUTHORS.
:license: CeCILL, see LICENSE for details.
"""
import numpy, logging, sys, re
from pyNN import errors, common, core, random, utility, recording, descriptions
from pyNN.space import Space
from pyNN.recording import files
from pyNN.random import RandomDistribution
from numpy import arccos, arcsin, arctan, arctan2, ceil, cos, cosh, e, exp, \
fabs, floor, fmod, hypot, ldexp, log, log10, modf, pi, power, \
sin, sinh, sqrt, tan, tanh, maximum, minimum
try:
import csa
haveCSA = True
except ImportError:
haveCSA = False
logger = logging.getLogger("PyNN")
def expand_distances(d_expression):
"""
Check if a distance expression contains at least one term d[x]. If yes, then
the distances are expanded and we assume the user has specified an
expression such as d[0] + d[2].
"""
regexpr = re.compile(r'.*d\[\d*\].*')
if regexpr.match(d_expression):
return True
return False
class ConnectionAttributeGenerator(object):
"""
Connection attributes, such as weights and delays, may be specified as:
- a single numerical value, in which case all connections have this value
- a numpy array of the same size as the number of connections
- a RandomDistribution object
- a function of the distance between the source and target cells
This class encapsulates all these different possibilities in order to
present a uniform interface.
"""
def __init__(self, source, local_mask, safe=True):
"""
Create a new %s.
source - something that may be used to obtain connection attribute values
local_mask - a boolean array indicating which of the post-synaptic cells
are on the local machine
safe - whether to check that values are within the appropriate range. These
checks can be slow, so safe=False allows you to turn them off once
you're certain your code is working correctly.
""" % self.__class__.__name__
self.source = source
self.local_mask = local_mask
self.safe = safe
if self.safe:
self.get = self.get_safe
if isinstance(self.source, numpy.ndarray):
self.source_iterator = iter(self.source)
def check(self, data):
"""
This method should be over-ridden by sub-classes.
"""
return data
def extract(self, N, distance_matrix=None, sub_mask=None):
"""
Return an array of values for a connection attribute.
N - number of values to be returned over the entire simulation. If
running a distributed simulation, the number returned on any given
node will be smaller.
distance_matrix - a DistanceMatrix object, used for calculating
distance-dependent attributes.
sub-mask - a sublist of the ids we want compute some values with. For
example in parallel, distances shoudl be computed only between a source
and local targets, since only connections with those targets are established.
Avoid useless computations...
"""
if isinstance(self.source, basestring):
assert distance_matrix is not None
if expand_distances(self.source):
d = distance_matrix.as_array(sub_mask, expand=True)
else:
d = distance_matrix.as_array(sub_mask)
values = eval(self.source)
return values
elif callable(self.source):
assert distance_matrix is not None
d = distance_matrix.as_array(sub_mask, expand=True)
values = self.source(d)
return values
elif numpy.isscalar(self.source):
if sub_mask is None:
values = numpy.ones((self.local_mask.sum(),))*self.source
else:
values = numpy.ones((len(sub_mask),))*self.source
return values # seems a bit wasteful to return an array of identical values
elif isinstance(self.source, RandomDistribution):
if sub_mask is None:
values = self.source.next(N, mask_local=self.local_mask)
else:
data = self.source.next(N, mask_local=self.local_mask)
if type(data) == numpy.float64:
data = numpy.array([data])
values = data[sub_mask]
return values
elif isinstance(self.source, numpy.ndarray):
if len(self.source.shape) == 2:
source_row = self.source_iterator.next()
values = source_row[self.local_mask]
elif len(self.source.shape) == 1: # for OneToOneConnector or AllToAllConnector used from or to only one Neuron
values = self.source[self.local_mask]
else:
raise Exception()
if sub_mask is not None:
values = values[sub_mask]
return values
else:
raise Exception("Invalid source")
def get_safe(self, N, distance_matrix=None, sub_mask=None):
return self.check(self.extract(N, distance_matrix, sub_mask))
def get(self, N, distance_matrix=None, sub_mask=None):
return self.extract(N, distance_matrix, sub_mask)
class WeightGenerator(ConnectionAttributeGenerator):
"""Generator for synaptic weights. %s""" % ConnectionAttributeGenerator.__doc__
def __init__(self, source, local_mask, projection, safe=True):
ConnectionAttributeGenerator.__init__(self, source, local_mask, safe)
self.projection = projection
self.is_conductance = common.is_conductance(projection.post.all_cells[0])
def check(self, weight):
if weight is None:
weight = common.DEFAULT_WEIGHT
if core.is_listlike(weight):
weight = numpy.array(weight)
nan_filter = (1-numpy.isnan(weight)).astype(bool) # weight arrays may contain NaN, which should be ignored
filtered_weight = weight[nan_filter]
all_negative = (filtered_weight<=0).all()
all_positive = (filtered_weight>=0).all()
if not (all_negative or all_positive):
raise errors.InvalidWeightError("Weights must be either all positive or all negative")
elif numpy.isscalar(weight):
all_positive = weight >= 0
all_negative = weight < 0
else:
raise Exception("Weight must be a number or a list/array of numbers.")
if self.is_conductance or self.projection.synapse_type == 'excitatory':
if not all_positive:
raise errors.InvalidWeightError("Weights must be positive for conductance-based and/or excitatory synapses")
elif self.is_conductance==False and self.projection.synapse_type == 'inhibitory':
if not all_negative:
raise errors.InvalidWeightError("Weights must be negative for current-based, inhibitory synapses")
else: # is_conductance is None. This happens if the cell does not exist on the current node.
logger.debug("Can't check weight, conductance status unknown.")
return weight
class DelayGenerator(ConnectionAttributeGenerator):
"""Generator for synaptic delays. %s""" % ConnectionAttributeGenerator.__doc__
def __init__(self, source, local_mask, safe=True):
ConnectionAttributeGenerator.__init__(self, source, local_mask, safe)
self.min_delay = common.get_min_delay()
self.max_delay = common.get_max_delay()
def check(self, delay):
all_negative = (delay<=self.max_delay).all()
all_positive = (delay>=self.min_delay).all()# If the delay is too small , we have to throw an error
if not (all_negative and all_positive):
raise errors.ConnectionError("delay (%s) is out of range [%s,%s]" % (delay, common.get_min_delay(), common.get_max_delay()))
return delay
class ProbaGenerator(ConnectionAttributeGenerator):
pass
class DistanceMatrix(object):
# should probably move to space module
def __init__(self, B, space, mask=None):
assert B.shape[0] == 3, B.shape
self.space = space
if mask is not None:
self.B = B[:,mask]
else:
self.B = B
def as_array(self, sub_mask=None, expand=False):
if self._distance_matrix is None and self.A is not None:
if sub_mask is None:
self._distance_matrix = self.space.distances(self.A, self.B, expand)
else:
self._distance_matrix = self.space.distances(self.A, self.B[:,sub_mask], expand)
if expand:
N = self._distance_matrix.shape[2]
self._distance_matrix = self._distance_matrix.reshape((3, N))
else:
self._distance_matrix = self._distance_matrix[0]
return self._distance_matrix
def set_source(self, A):
assert A.shape == (3,), A.shape
self.A = A
self._distance_matrix = None
class Connector(object):
def __init__(self, weights=0.0, delays=None, space=Space(), safe=True, verbose=False):
self.weights = weights
self.space = space
self.safe = safe
self.verbose = verbose
min_delay = common.get_min_delay()
if delays is None:
self.delays = min_delay
else:
if core.is_listlike(delays):
if min(delays) < min_delay:
raise errors.ConnectionError("smallest delay (%g) is smaller than minimum delay (%g)" % (min(delays), min_delay))
elif not (isinstance(delays, basestring) or isinstance(delays, RandomDistribution)):
if delays < min_delay:
raise errors.ConnectionError("delay (%g) is smaller than minimum delay (%g)" % (delays, min_delay))
self.delays = delays
def connect(self, projection):
raise NotImplementedError()
def progressbar(self, N):
self.prog = utility.ProgressBar(0, N, 20, mode='fixed')
def progression(self, count):
self.prog.update_amount(count)
if self.verbose and common.rank() == 0:
print self.prog, "\r",
sys.stdout.flush()
def get_parameters(self):
P = {}
for name in self.parameter_names:
P[name] = getattr(self, name)
return P
def describe(self, template='connector_default.txt', engine='default'):
"""
Returns a human-readable description of the connection method.
The output may be customized by specifying a different template
togther with an associated template engine (see ``pyNN.descriptions``).
If template is None, then a dictionary containing the template context
will be returned.
"""
context = {'name': self.__class__.__name__,
'parameters': self.get_parameters(),
'weights': self.weights,
'delays': self.delays}
return descriptions.render(engine, template, context)
class ProbabilisticConnector(Connector):
def __init__(self, projection, weights=0.0, delays=None,
allow_self_connections=True, space=Space(), safe=True):
Connector.__init__(self, weights, delays, space, safe)
if isinstance(projection.rng, random.NativeRNG):
raise Exception("Use of NativeRNG not implemented.")
else:
self.rng = projection.rng
self.local = projection.post._mask_local
self.N = projection.post.size
self.weights_generator = WeightGenerator(weights, self.local, projection, safe)
self.delays_generator = DelayGenerator(delays, self.local, safe)
self.probas_generator = ProbaGenerator(RandomDistribution('uniform', (0,1), rng=self.rng), self.local)
self._distance_matrix = None
self.projection = projection
self.candidates = projection.post.local_cells
self.size = self.local.sum()
self.allow_self_connections = allow_self_connections
@property
def distance_matrix(self):
"""
We want to avoid calculating positions if it is not necessary, so we
delay it until the distance matrix is actually used.
"""
if self._distance_matrix is None:
self._distance_matrix = DistanceMatrix(self.projection.post.positions, self.space, self.local)
return self._distance_matrix
def _probabilistic_connect(self, src, p, n_connections=None):
"""
Connect-up a Projection with connection probability p, where p may be either
a float 0<=p<=1, or a dict containing a float array for each pre-synaptic
cell, the array containing the connection probabilities for all the local
targets of that pre-synaptic cell.
"""
if numpy.isscalar(p) and p == 1:
precreate = numpy.arange(self.size)
else:
rarr = self.probas_generator.get(self.N)
if not core.is_listlike(rarr) and numpy.isscalar(rarr): # if N=1, rarr will be a single number
rarr = numpy.array([rarr])
precreate = numpy.where(rarr < p)[0]
self.distance_matrix.set_source(src.position)
if not self.allow_self_connections and self.projection.pre == self.projection.post:
idx_src = numpy.where(self.candidates == src)
if len(idx_src) > 0:
i = numpy.where(precreate == idx_src[0])
if len(i) > 0:
precreate = numpy.delete(precreate, i[0])
if (n_connections is not None) and (len(precreate) > 0):
create = numpy.array([], int)
while len(create) < n_connections: # if the number of requested cells is larger than the size of the
## presynaptic population, we allow multiple connections for a given cell
create = numpy.concatenate((create, self.projection.rng.permutation(precreate)))
create = create[:n_connections]
else:
create = precreate
targets = self.candidates[create]
weights = self.weights_generator.get(self.N, self.distance_matrix, create)
delays = self.delays_generator.get(self.N, self.distance_matrix, create)
if len(targets) > 0:
self.projection.connection_manager.connect(src, targets.tolist(), weights, delays)
class AllToAllConnector(Connector):
"""
Connects all cells in the presynaptic population to all cells in the
postsynaptic population.
"""
parameter_names = ('allow_self_connections',)
def __init__(self, allow_self_connections=True, weights=0.0, delays=None, space=Space(), safe=True, verbose=False):
"""
Create a new connector.
`allow_self_connections` -- if the connector is used to connect a
Population to itself, this flag determines whether a neuron is
allowed to connect to itself, or only to other neurons in the
Population.
`weights` -- may either be a float, a RandomDistribution object, a list/
1D array with at least as many items as connections to be
created. Units nA.
`delays` -- as `weights`. If `None`, all synaptic delays will be set
to the global minimum delay.
`space` -- a `Space` object, needed if you wish to specify distance-
dependent weights or delays
"""
Connector.__init__(self, weights, delays, space, safe, verbose)
assert isinstance(allow_self_connections, bool)
self.allow_self_connections = allow_self_connections
def connect(self, projection):
connector = ProbabilisticConnector(projection, self.weights, self.delays, self.allow_self_connections, self.space, safe=self.safe)
self.progressbar(len(projection.pre))
for count, src in enumerate(projection.pre.all()):
connector._probabilistic_connect(src, 1)
self.progression(count)
class FixedProbabilityConnector(Connector):
"""
For each pair of pre-post cells, the connection probability is constant.
"""
parameter_names = ('allow_self_connections', 'p_connect')
def __init__(self, p_connect, allow_self_connections=True, weights=0.0,
delays=None, space=Space(), safe=True, verbose=False):
"""
Create a new connector.
`p_connect` -- a float between zero and one. Each potential connection
is created with this probability.
`allow_self_connections` -- if the connector is used to connect a
Population to itself, this flag determines whether a neuron is
allowed to connect to itself, or only to other neurons in the
Population.
`weights` -- may either be a float, a RandomDistribution object, a list/
1D array with at least as many items as connections to be
created. Units nA.
`delays` -- as `weights`. If `None`, all synaptic delays will be set
to the global minimum delay.
`space` -- a `Space` object, needed if you wish to specify distance-
dependent weights or delays
"""
Connector.__init__(self, weights, delays, space, safe, verbose)
assert isinstance(allow_self_connections, bool)
self.allow_self_connections = allow_self_connections
self.p_connect = float(p_connect)
assert 0 <= self.p_connect
def connect(self, projection):
#assert projection.rng.parallel_safe
connector = ProbabilisticConnector(projection, self.weights, self.delays,
self.allow_self_connections, self.space,
safe=self.safe)
self.progressbar(len(projection.pre))
for count, src in enumerate(projection.pre.all()):
connector._probabilistic_connect(src, self.p_connect)
self.progression(count)
class DistanceDependentProbabilityConnector(Connector):
"""
For each pair of pre-post cells, the connection probability depends on distance.
"""
parameter_names = ('allow_self_connections', 'd_expression')
def __init__(self, d_expression, allow_self_connections=True,
weights=0.0, delays=None, space=Space(), safe=True, verbose=False, n_connections=None):
"""
Create a new connector.
`d_expression` -- the right-hand side of a valid python expression for
probability, involving 'd', e.g. "exp(-abs(d))", or "d<3"
`n_connections` -- The number of efferent synaptic connections per neuron.
`space` -- a Space object.
`weights` -- may either be a float, a RandomDistribution object, a list/
1D array with at least as many items as connections to be
created, or a distance expression as for `d_expression`. Units nA.
`delays` -- as `weights`. If `None`, all synaptic delays will be set
to the global minimum delay.
"""
Connector.__init__(self, weights, delays, space, safe, verbose)
assert isinstance(d_expression, str)
try:
if not expand_distances(d_expression):
d = 0; assert 0 <= eval(d_expression), eval(d_expression)
d = 1e12; assert 0 <= eval(d_expression), eval(d_expression)
except ZeroDivisionError, err:
raise ZeroDivisionError("Error in the distance expression %s. %s" % (d_expression, err))
self.d_expression = d_expression
assert isinstance(allow_self_connections, bool)
self.allow_self_connections = allow_self_connections
self.n_connections = n_connections
def connect(self, projection):
"""Connect-up a Projection."""
connector = ProbabilisticConnector(projection, self.weights, self.delays, self.allow_self_connections, self.space, safe=self.safe)
proba_generator = ProbaGenerator(self.d_expression, connector.local)
self.progressbar(len(projection.pre))
if (common.num_processes() > 1) and (self.n_connections is not None):
raise Exception("n_connections not implemented yet for this connector in parallel !")
for count, src in enumerate(projection.pre.all()):
connector.distance_matrix.set_source(src.position)
proba = proba_generator.get(connector.N, connector.distance_matrix)
if proba.dtype == 'bool':
proba = proba.astype(float)
connector._probabilistic_connect(src, proba, self.n_connections)
self.progression(count)
class FromListConnector(Connector):
"""
Make connections according to a list.
"""
parameter_names = ('conn_list',)
def __init__(self, conn_list, safe=True, verbose=False):
"""
Create a new connector.
`conn_list` -- a list of tuples, one tuple for each connection. Each
tuple should contain:
(pre_idx, post_idx, weight, delay)
where pre_idx is the index (i.e. order in the Population,
not the ID) of the presynaptic neuron, and post_idx is
the index of the postsynaptic neuron.
"""
# needs extending for dynamic synapses.
Connector.__init__(self, 0., common.get_min_delay(), safe=safe, verbose=verbose)
self.conn_list = numpy.array(conn_list)
def connect(self, projection):
"""Connect-up a Projection."""
idx = numpy.argsort(self.conn_list[:, 0])
self.sources = numpy.unique(self.conn_list[:,0]).astype(int)
self.candidates = projection.post.local_cells
self.conn_list = self.conn_list[idx]
self.progressbar(len(self.sources))
count = 0
left = numpy.searchsorted(self.conn_list[:,0], self.sources, 'left')
right = numpy.searchsorted(self.conn_list[:,0], self.sources, 'right')
#tests = "|".join(['(tgts == %d)' %id for id in self.candidates])
for src, l, r in zip(self.sources, left, right):
targets = self.conn_list[l:r, 1].astype(int)
weights = self.conn_list[l:r, 2]
delays = self.conn_list[l:r, 3]
try:
src = projection.pre.all_cells[src]
except IndexError:
raise errors.ConnectionError("invalid source index %s" % src)
try:
tgts = projection.post.all_cells[targets]
except IndexError:
raise errors.ConnectionError("invalid target index or indices")
## We need to exclude the non local cells. Fastidious, need maybe
## to use a convergent_connect method, instead of a divergent_connect one
#idx = eval(tests)
#projection.connection_manager.connect(src, tgts[idx].tolist(), weights[idx], delays[idx])
projection.connection_manager.connect(src, tgts.tolist(), weights, delays)
self.progression(count)
count += 1
class FromFileConnector(FromListConnector):
"""
Make connections according to a list read from a file.
"""
parameter_names = ('filename', 'distributed')
def __init__(self, file, distributed=False, safe=True, verbose=False):
"""
Create a new connector.
`file` -- file object containing a list of connections, in
the format required by `FromListConnector`.
`distributed` -- if this is True, then each node will read connections
from a file called `filename.x`, where `x` is the MPI
rank. This speeds up loading connections for
distributed simulations.
"""
Connector.__init__(self, 0., common.get_min_delay(), safe=safe, verbose=verbose)
if isinstance(file, basestring):
file = files.StandardTextFile(file, mode='r')
self.file = file
self.distributed = distributed
def connect(self, projection):
"""Connect-up a Projection."""
if self.distributed:
self.file.rename("%s.%d" % (self.file.name, common.rank()))
self.conn_list = self.file.read()
FromListConnector.connect(self, projection)
class FixedNumberPostConnector(Connector):
"""
Each pre-synaptic neuron is connected to exactly n post-synaptic neurons
chosen at random.
If n is less than the size of the post-synaptic population, there are no
multiple connections, i.e., no instances of the same pair of neurons being
multiply connected. If n is greater than the size of the post-synaptic
population, all possible single connections are made before starting to add
duplicate connections.
"""
parameter_names = ('allow_self_connections', 'n')
def __init__(self, n, allow_self_connections=True, weights=0.0, delays=None, space=Space(), safe=True, verbose=False):
"""
Create a new connector.
`n` -- either a positive integer, or a `RandomDistribution` that produces
positive integers. If `n` is a `RandomDistribution`, then the
number of post-synaptic neurons is drawn from this distribution
for each pre-synaptic neuron.
`allow_self_connections` -- if the connector is used to connect a
Population to itself, this flag determines whether a neuron is
allowed to connect to itself, or only to other neurons in the
Population.
`weights` -- may either be a float, a RandomDistribution object, a list/
1D array with at least as many items as connections to be
created. Units nA.
`delays` -- as `weights`. If `None`, all synaptic delays will be set
to the global minimum delay.
"""
Connector.__init__(self, weights, delays, space, safe, verbose)
assert isinstance(allow_self_connections, bool)
self.allow_self_connections = allow_self_connections
if isinstance(n, int):
self.n = n
assert n >= 0
elif isinstance(n, random.RandomDistribution):
self.rand_distr = n
# weak check that the random distribution is ok
assert numpy.all(numpy.array(n.next(100)) >= 0), "the random distribution produces negative numbers"
else:
raise Exception("n must be an integer or a RandomDistribution object")
def connect(self, projection):
"""Connect-up a Projection."""
local = numpy.ones(len(projection.post), bool)
weights_generator = WeightGenerator(self.weights, local, projection, self.safe)
delays_generator = DelayGenerator(self.delays, local, self.safe)
distance_matrix = DistanceMatrix(projection.post.positions, self.space)
candidates = projection.post.all_cells
size = len(projection.post)
self.progressbar(len(projection.pre))
if isinstance(projection.rng, random.NativeRNG):
raise Exception("Use of NativeRNG not implemented.")
for count, src in enumerate(projection.pre.all()):
# pick n neurons at random
if hasattr(self, 'rand_distr'):
n = self.rand_distr.next()
else:
n = self.n
idx = numpy.arange(size)
if not self.allow_self_connections and projection.pre == projection.post:
i = numpy.where(candidates == src)[0]
idx = numpy.delete(idx, i)
create = numpy.array([])
while len(create) < n: # if the number of requested cells is larger than the size of the
# postsynaptic population, we allow multiple connections for a given cell
create = numpy.concatenate((create, projection.rng.permutation(idx)[:n]))
distance_matrix.set_source(src.position)
create = create[:n].astype(int)
targets = candidates[create]
weights = weights_generator.get(n, distance_matrix, create)
delays = delays_generator.get(n, distance_matrix, create)
if len(targets) > 0:
projection.connection_manager.connect(src, targets.tolist(), weights, delays)
self.progression(count)
class FixedNumberPreConnector(Connector):
"""
Each post-synaptic neuron is connected to exactly n pre-synaptic neurons
chosen at random.
If n is less than the size of the pre-synaptic population, there are no
multiple connections, i.e., no instances of the same pair of neurons being
multiply connected. If n is greater than the size of the pre-synaptic
population, all possible single connections are made before starting to add
duplicate connections.
"""
parameter_names = ('allow_self_connections', 'n')
def __init__(self, n, allow_self_connections=True, weights=0.0, delays=None, space=Space(), safe=True, verbose=False):
"""
Create a new connector.
`n` -- either a positive integer, or a `RandomDistribution` that produces
positive integers. If `n` is a `RandomDistribution`, then the
number of pre-synaptic neurons is drawn from this distribution
for each post-synaptic neuron.
`allow_self_connections` -- if the connector is used to connect a
Population to itself, this flag determines whether a neuron is
allowed to connect to itself, or only to other neurons in the
Population.
`weights` -- may either be a float, a RandomDistribution object, a list/
1D array with at least as many items as connections to be
created. Units nA.
`delays` -- as `weights`. If `None`, all synaptic delays will be set
to the global minimum delay.
"""
Connector.__init__(self, weights, delays, space, safe, verbose)
assert isinstance(allow_self_connections, bool)
self.allow_self_connections = allow_self_connections
if isinstance(n, int):
self.n = n
assert n >= 0
elif isinstance(n, random.RandomDistribution):
self.rand_distr = n
# weak check that the random distribution is ok
assert numpy.all(numpy.array(n.next(100)) >= 0), "the random distribution produces negative numbers"
else:
raise Exception("n must be an integer or a RandomDistribution object")
def connect(self, projection):
"""Connect-up a Projection."""
local = numpy.ones(len(projection.pre), bool)
weights_generator = WeightGenerator(self.weights, local, projection, self.safe)
delays_generator = DelayGenerator(self.delays, local, self.safe)
distance_matrix = DistanceMatrix(projection.pre.positions, self.space)
candidates = projection.pre.all_cells
size = len(projection.pre)
self.progressbar(len(projection.post.local_cells))
if isinstance(projection.rng, random.NativeRNG):
raise Exception("Warning: use of NativeRNG not implemented.")
for count, tgt in enumerate(projection.post.local_cells):
# pick n neurons at random
if hasattr(self, 'rand_distr'):
n = self.rand_distr.next()
else:
n = self.n
idx = numpy.arange(size)
if not self.allow_self_connections and projection.pre == projection.post:
i = numpy.where(candidates == tgt)[0]
idx = numpy.delete(idx, i)
create = numpy.array([])
while len(create) < n: # if the number of requested cells is larger than the size of the
# presynaptic population, we allow multiple connections for a given cell
create = numpy.concatenate((create, projection.rng.permutation(idx)[:n]))
distance_matrix.set_source(tgt.position)
create = create[:n].astype(int)
sources = candidates[create]
weights = weights_generator.get(n, distance_matrix, create)
delays = delays_generator.get(n, distance_matrix, create)
for src, w, d in zip(sources, weights, delays):
projection.connection_manager.connect(src, tgt, w, d)
self.progression(count)
class OneToOneConnector(Connector):
"""
Where the pre- and postsynaptic populations have the same size, connect
cell i in the presynaptic population to cell i in the postsynaptic
population for all i.
"""
parameter_names = tuple()
def __init__(self, weights=0.0, delays=None, space=Space(), safe=True, verbose=False):
"""
Create a new connector.
`weights` -- may either be a float, a RandomDistribution object, a list/
1D array with at least as many items as connections to be
created. Units nA.
`delays` -- as `weights`. If `None`, all synaptic delays will be set
to the global minimum delay.
"""
Connector.__init__(self, weights, delays, space, verbose)
self.space = space
self.safe = safe
def connect(self, projection):
"""Connect-up a Projection."""
if projection.pre.size == projection.post.size:
N = projection.post.size
local = projection.post._mask_local
if isinstance(self.weights, basestring) or isinstance(self.delays, basestring):
raise Exception('Expression for weights or delays is not supported for OneToOneConnector !')
weights_generator = WeightGenerator(self.weights, local, projection, self.safe)
delays_generator = DelayGenerator(self.delays, local, self.safe)
weights = weights_generator.get(N)
delays = delays_generator.get(N)
self.progressbar(len(projection.post.local_cells))
count = 0
create = numpy.arange(0, N)[local]
sources = projection.pre.all_cells[create]
for tgt, src, w, d in zip(projection.post.local_cells, sources, weights, delays):
# the float is in case the values are of type numpy.float64, which NEST chokes on
projection.connection_manager.connect(src, [tgt], [float(w)], [float(d)])
self.progression(count)
count += 1
else:
raise errors.InvalidDimensionsError("OneToOneConnector does not support presynaptic and postsynaptic Populations of different sizes.")
class SmallWorldConnector(Connector):
"""
For each pair of pre-post cells, the connection probability depends on distance.
"""
parameter_names = ('allow_self_connections', 'degree', 'rewiring')
def __init__(self, degree, rewiring, allow_self_connections=True,
weights=0.0, delays=None, space=Space(), safe=True, verbose=False, n_connections=None):
"""
Create a new connector.
`degree` -- the region lenght where nodes will be connected locally
`rewiring` -- the probability of rewiring each eadges
`space` -- a Space object.
`allow_self_connections` -- if the connector is used to connect a
Population to itself, this flag determines whether a neuron is
allowed to connect to itself, or only to other neurons in the
Population.
`n_connections` -- The number of efferent synaptic connections per neuron.
`weights` -- may either be a float, a RandomDistribution object, a list/
1D array with at least as many items as connections to be
created, or a DistanceDependence object. Units nA.
`delays` -- as `weights`. If `None`, all synaptic delays will be set
to the global minimum delay.
"""
Connector.__init__(self, weights, delays, space, safe, verbose)
assert 0 <= rewiring <= 1
assert isinstance(allow_self_connections, bool)
self.rewiring = rewiring
self.d_expression = "d < %g" %degree
self.allow_self_connections = allow_self_connections
self.n_connections = n_connections
def _smallworld_connect(self, src, p, n_connections=None):
"""
Connect-up a Projection with connection probability p, where p may be either
a float 0<=p<=1, or a dict containing a float array for each pre-synaptic
cell, the array containing the connection probabilities for all the local
targets of that pre-synaptic cell.
"""
rarr = self.probas_generator.get(self.N)
if not core.is_listlike(rarr) and numpy.isscalar(rarr): # if N=1, rarr will be a single number
rarr = numpy.array([rarr])
precreate = numpy.where(rarr < p)[0]
self.distance_matrix.set_source(src.position)
if not self.allow_self_connections and self.projection.pre == self.projection.post:
i = numpy.where(self.candidates == src)[0]
precreate = numpy.delete(precreate, i)
idx = numpy.arange(0, self.size)
if not self.allow_self_connections and self.projection.pre == self.projection.post:
i = numpy.where(self.candidates == src)[0]
idx = numpy.delete(idx, i)
rarr = self.probas_generator.get(self.N)[precreate]
rewired = numpy.where(rarr < self.rewiring)[0]
N = len(rewired)
if N > 0:
new_idx = (len(idx)-1) * self.probas_generator.get(self.N)[precreate]
precreate[rewired] = idx[new_idx.astype(int)]
if (n_connections is not None) and (len(precreate) > 0):
create = numpy.array([], int)
while len(create) < n_connections: # if the number of requested cells is larger than the size of the
## presynaptic population, we allow multiple connections for a given cell
create = numpy.concatenate((create, self.projection.rng.permutation(precreate)))
create = create[:n_connections]
else:
create = precreate
targets = self.candidates[create]
weights = self.weights_generator.get(self.N, self.distance_matrix, create)
delays = self.delays_generator.get(self.N, self.distance_matrix, create)
if len(targets) > 0:
self.projection.connection_manager.connect(src, targets.tolist(), weights, delays)
def connect(self, projection):
"""Connect-up a Projection."""
local = numpy.ones(len(projection.post), bool)
self.N = projection.post.size
if isinstance(projection.rng, random.NativeRNG):
raise Exception("Use of NativeRNG not implemented.")
else:
self.rng = projection.rng
self.weights_generator = WeightGenerator(self.weights, local, projection, self.safe)
self.delays_generator = DelayGenerator(self.delays, local, self.safe)
self.probas_generator = ProbaGenerator(RandomDistribution('uniform',(0,1), rng=self.rng), local)
self.distance_matrix = DistanceMatrix(projection.post.positions, self.space, local)
self.projection = projection
self.candidates = projection.post.all_cells
self.size = len(projection.post)
self.progressbar(len(projection.pre))
proba_generator = ProbaGenerator(self.d_expression, local)
for count, src in enumerate(projection.pre.all()):
self.distance_matrix.set_source(src.position)
proba = proba_generator.get(self.N, self.distance_matrix).astype(float)
self._smallworld_connect(src, proba, self.n_connections)
self.progression(count)
class CSAConnector(Connector):
parameter_names = ('cset',)
if haveCSA:
def __init__ (self, cset, weights=None, delays=None, safe=True, verbose=False):
"""
"""
min_delay = common.get_min_delay()
Connector.__init__(self, None, None, safe=safe, verbose=verbose)
self.cset = cset
if csa.arity(cset) == 0:
#assert weights is not None and delays is not None, \
# 'must specify weights and delays in addition to a CSA mask'
self.weights = weights
if weights is None:
self.weights = common.DEFAULT_WEIGHT
self.delays = delays
if delays is None:
self.delays = common.get_min_delay()
else:
assert csa.arity(cset) == 2, 'must specify mask or connection-set with arity 2'
assert weights is None and delays is None, \
"weights or delays specified both in connection-set and as CSAConnector argument"
else:
def __init__ (self, cset, safe=True, verbose=False):
raise RuntimeError, "CSAConnector not available---couldn't import csa module"
@staticmethod
def isConstant (x):
return isinstance (x, (int, float))
@staticmethod
def constantIterator (x):
while True:
yield x
def connect(self, projection):
"""Connect-up a Projection."""
# Cut out finite part
c = csa.cross((0, projection.pre.size-1), (0, projection.post.size-1)) * self.cset
if csa.arity(self.cset) == 2:
# Connection-set with arity 2
for (i, j, weight, delay) in c:
projection.connection_manager.connect (projection.pre[i], [projection.post[j]], weight, delay)
elif CSAConnector.isConstant (self.weights) \
and CSAConnector.isConstant (self.delays):
# Mask with constant weights and delays
for (i, j) in c:
projection.connection_manager.connect (projection.pre[i], [projection.post[j]], self.weights, self.delays)
else:
# Mask with weights and/or delays iterable
weights = self.weights
if CSAConnector.isConstant (weights):
weights = CSAConnector.constantIterator (weights)
delays = self.delays
if CSAConnector.isConstant (delays):
delays = CSAConnector.constantIterator (delays)
for (i, j), weight, delay in zip (c, weights, delays):
projection.connection_manager.connect (projection.pre[i], [projection.post[j]], weight, delay)
|