__init__.py 19.8 KB
Newer Older
Tiago Peixoto's avatar
Tiago Peixoto committed
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
#! /usr/bin/env python
# graph_tool.py -- a general graph manipulation python module
#
# Copyright (C) 2007 Tiago de Paula Peixoto <tiago@forked.de>
#
# This program is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program.  If not, see <http://www.gnu.org/licenses/>.

19
"""
20 21
``graph_tool.centrality`` - Centrality measures
-----------------------------------------------
22 23

This module includes centrality-related algorithms.
24 25 26 27 28 29 30 31 32 33 34 35 36 37 38

Summary
+++++++

.. autosummary::
   :nosignatures:

   pagerank
   betweenness
   central_point_dominance
   eigentrust
   absolute_trust

Contents
++++++++
39 40
"""

Tiago Peixoto's avatar
Tiago Peixoto committed
41 42 43 44
from .. dl_import import dl_import
dl_import("import libgraph_tool_centrality")

from .. core import _prop
45
import sys, numpy
Tiago Peixoto's avatar
Tiago Peixoto committed
46 47 48 49

__all__ = ["pagerank", "betweenness", "central_point_dominance", "eigentrust",
           "absolute_trust"]

50
def pagerank(g, damping=0.8, prop=None, epslon=1e-6, max_iter=None,
Tiago Peixoto's avatar
Tiago Peixoto committed
51
             ret_iter=False):
52 53 54 55 56
    r"""
    Calculate the PageRank of each vertex.

    Parameters
    ----------
57
    g : :class:`~graph_tool.Graph`
58 59 60
        Graph to be used.
    damping : float, optional (default: 0.8)
        Damping factor.
61
    prop : :class:`~graph_tool.PropertyMap`, optional (default: None)
62 63 64 65 66 67 68 69 70 71 72
        Vertex property map to store the PageRank values.
    epslon : float, optional (default: 1e-6)
        Convergence condition. The iteration will stop if the total delta of all
        vertices are below this value.
    max_iter : int, optional (default: None)
        If supplied, this will limit the total number of iterations.
    ret_iter : bool, optional (default: False)
        If true, the total number of iterations is also returned.

    Returns
    -------
73 74
    pagerank : :class:`~graph_tool.PropertyMap`
        A vertex property map containing the PageRank values.
75 76 77 78 79 80 81 82 83

    See Also
    --------
    betweenness: betweenness centrality
    eigentrust: eigentrust centrality
    absolute_trust: absolute trust centrality

    Notes
    -----
84
    The value of PageRank [pagerank-wikipedia]_ of vertex v :math:`PR(v)` is
85 86 87
    given interactively by the relation:

    .. math::
88 89

        PR(v) = \frac{1-d}{N} + d \sum_{w \in \Gamma^{-}(v)}
90
                \frac{PR (w)}{d^{+}(w)}
91 92 93 94 95 96 97 98 99 100 101 102

    where :math:`\Gamma^{-}(v)` are the in-neighbours of v, :math:`d^{+}(w)` is
    the out-degree of w, and d is a damping factor.

    The implemented algorithm progressively iterates the above condition, until
    it no longer changes, according to the parameter epslon. It has a
    topology-dependent running time.

    If enabled during compilation, this algorithm runs in parallel.

    Examples
    --------
103 104
    >>> from numpy.random import poisson, seed
    >>> seed(42)
105
    >>> g = gt.random_graph(100, lambda: (poisson(3), poisson(3)))
106
    >>> pr = gt.pagerank(g)
107
    >>> print pr.a
Tiago Peixoto's avatar
Tiago Peixoto committed
108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124
    [ 0.63876901  1.13528868  0.31465963  0.55855277  0.2         0.75605741
      0.42628689  0.53066254  0.55004112  0.91717076  0.71164749  0.32015438
      0.67275227  1.08207389  1.14412231  0.9049167   1.32002     1.4692142
      0.76549771  0.71510277  0.23732927  0.40844911  0.2         0.27912876
      0.71309781  0.32015438  1.3376236   0.31352887  0.59346569  0.33381039
      0.67300081  0.73318264  0.65812653  0.73409673  0.93051993  0.83241145
      1.59816568  0.43979363  0.2512247   1.15663357  0.2         0.35977148
      0.72182022  1.01267711  0.76304859  0.49247376  0.49384283  1.8436647
      0.64312224  1.00778243  0.62287633  1.15215387  0.56176895  0.7166227
      0.56506109  0.67104337  0.95570565  0.27996953  0.79975983  0.33631497
      1.09471419  0.33631497  0.2512247   2.09126732  0.68157485  0.2
      0.37140185  0.65619459  1.27370737  0.48383225  1.36125161  0.2
      0.78300573  1.03427279  0.56904755  1.66077917  1.73302035  0.28749261
      0.83143045  1.04969728  0.70090048  0.55991433  0.68440994  0.2
      0.34018009  0.45485484  0.28        1.2015438   2.11850885  1.24990775
      0.59914308  0.59989185  0.73535564  0.78168417  0.55390281  0.38627667
      1.42274704  0.51105348  0.92550979  1.27968065]
125 126 127

    References
    ----------
128 129
    .. [pagerank-wikipedia] http://en.wikipedia.org/wiki/Pagerank
    .. [lawrence-pagerank-1998] P. Lawrence, B. Sergey, M. Rajeev, W. Terry,
130
       "The pagerank citation ranking: Bringing order to the web", Technical
131 132 133 134 135
       report, Stanford University, 1998
    """

    if max_iter == None:
        max_iter = 0
Tiago Peixoto's avatar
Tiago Peixoto committed
136 137 138 139 140 141 142 143 144 145
    if prop == None:
        prop = g.new_vertex_property("double")
    ic = libgraph_tool_centrality.\
            get_pagerank(g._Graph__graph, _prop("v", g, prop), damping, epslon,
                         max_iter)
    if ret_iter:
        return prop, ic
    else:
        return prop

146 147 148 149 150 151
def betweenness(g, vprop=None, eprop=None, weight=None, norm=True):
    r"""
    Calculate the betweenness centrality for each vertex and edge.

    Parameters
    ----------
152
    g : :class:`~graph_tool.Graph`
153
        Graph to be used.
154
    vprop : :class:`~graph_tool.PropertyMap`, optional (default: None)
155
        Vertex property map to store the vertex betweenness values.
156
    eprop : :class:`~graph_tool.PropertyMap`, optional (default: None)
157
        Edge property map to store the edge betweenness values.
158
    weight : :class:`~graph_tool.PropertyMap`, optional (default: None)
159 160 161 162 163 164
        Edge property map corresponding to the weight value of each edge.
    norm : bool, optional (default: True)
        Whether or not the betweenness values should be normalized.

    Returns
    -------
165 166 167 168
    vertex_betweenness : A vertex property map with the vertex betweenness
                         values.
    edge_betweenness : An edge property map with the edge betweenness
                       values.
169 170 171 172 173 174 175 176 177 178 179 180

    See Also
    --------
    central_point_dominance: central point dominance of the graph
    pagerank: PageRank centrality
    eigentrust: eigentrust centrality
    absolute_trust: absolute trust centrality

    Notes
    -----
    Betweenness centrality of a vertex :math:`C_B(v)` is defined as,

181 182
    .. math::

183 184 185 186 187 188 189 190 191
        C_B(v)= \sum_{s \neq v \neq t \in V \atop s \neq t}
                \frac{\sigma_{st}(v)}{\sigma_{st}}

    where :math:`\sigma_{st}` is the number of shortest geodesic paths from s to
    t, and :math:`\sigma_{st}(v)` is the number of shortest geodesic paths from
    s to t that pass through a vertex v.  This may be normalised by dividing
    through the number of pairs of vertices not including v, which is
    :math:`(n-1)(n-2)/2`.

192
    The algorithm used here is defined in [brandes-faster-2001]_, and has a
193 194 195 196 197 198 199
    complexity of :math:`O(VE)` for unweighted graphs and :math:`O(VE + V(V+E)
    \log V)` for weighted graphs. The space complexity is :math:`O(VE)`.

    If enabled during compilation, this algorithm runs in parallel.

    Examples
    --------
200 201
    >>> from numpy.random import poisson, seed
    >>> seed(42)
202
    >>> g = gt.random_graph(100, lambda: (poisson(3), poisson(3)))
203
    >>> vb, eb = gt.betweenness(g)
204
    >>> print vb.a
Tiago Peixoto's avatar
Tiago Peixoto committed
205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221
    [ 0.03395047  0.07911989  0.00702948  0.02337119  0.          0.02930099
      0.01684377  0.02558675  0.03440095  0.02886187  0.03124262  0.00975953
      0.01307953  0.03938858  0.07266505  0.01313647  0.          0.06450598
      0.0575418   0.00525468  0.00466089  0.01803829  0.          0.00050161
      0.0085034   0.02362432  0.05620574  0.00097157  0.04006816  0.01301474
      0.02154916  0.          0.06009194  0.02780363  0.08963522  0.04049657
      0.06993559  0.02082698  0.00288318  0.03264322  0.          0.03641759
      0.01083859  0.03750864  0.04079359  0.02092599  0.          0.02153655
      0.          0.05674631  0.03861911  0.05473282  0.00904367  0.03249097
      0.00894043  0.0192741   0.03379204  0.02125998  0.0018321   0.0013495
      0.0336502   0.0210088   0.00125318  0.0489189   0.05254974  0.
      0.00432189  0.04866168  0.06444727  0.02508525  0.02533085  0.
      0.05308703  0.02539854  0.02270809  0.044889    0.04766016  0.0086368
      0.01501699  0.          0.03107868  0.0054221   0.          0.
      0.00596081  0.01183977  0.00159761  0.11435876  0.03988501  0.05128991
      0.04558135  0.02303469  0.05092032  0.04700221  0.00927644  0.00841903
      0.          0.03243633  0.04514374  0.05170213]
222 223 224

    References
    ----------
225 226
    .. [betweenness-wikipedia] http://en.wikipedia.org/wiki/Centrality#Betweenness_centrality
    .. [brandes-faster-2001] U. Brandes, "A faster algorithm for betweenness
227 228
       centrality",  Journal of Mathematical Sociology, 2001
    """
Tiago Peixoto's avatar
Tiago Peixoto committed
229 230 231 232 233 234 235 236 237 238 239 240 241 242
    if vprop == None:
        vprop = g.new_vertex_property("double")
    if eprop == None:
        eprop = g.new_edge_property("double")
    if weight != None and weight.value_type() != eprop.value_type():
        nw = g.new_edge_property(eprop.value_type())
        g.copy_property(weight, nw)
        weight = nw
    libgraph_tool_centrality.\
            get_betweenness(g._Graph__graph, _prop("e", g, weight),
                            _prop("e", g, eprop), _prop("v", g, vprop), norm)
    return vprop, eprop

def central_point_dominance(g, betweenness):
243 244 245 246 247 248
    r"""
    Calculate the central point dominance of the graph, given the betweenness
    centrality of each vertex.

    Parameters
    ----------
249
    g : :class:`~graph_tool.Graph`
250
        Graph to be used.
251
    betweenness : :class:`~graph_tool.PropertyMap`
252 253 254 255 256
        Vertex property map with the betweenness centrality values. The values
        must be normalized.

    Returns
    -------
257 258
    cp : float
        The central point dominance.
259 260 261 262 263 264 265 266

    See Also
    --------
    betweenness: betweenness centrality

    Notes
    -----
    Let :math:`v^*` be the vertex with the largest relative betweenness
267
    centrality; then, the central point dominance [freeman-set-1977]_ is defined
268 269
    as:

270 271
    .. math::

272 273 274 275 276 277 278 279 280
        C'_B = \frac{1}{|V|-1} \sum_{v} C_B(v^*) - C_B(v)

    where :math:`C_B(v)` is the normalized betweenness centrality of vertex
    v. The value of :math:`C_B` lies in the range [0,1].

    The algorithm has a complexity of :math:`O(V)`.

    Examples
    --------
281 282
    >>> from numpy.random import poisson, seed
    >>> seed(42)
283
    >>> g = gt.random_graph(100, lambda: (poisson(3), poisson(3)))
284 285
    >>> vb, eb = gt.betweenness(g)
    >>> print gt.central_point_dominance(g, vb)
Tiago Peixoto's avatar
Tiago Peixoto committed
286
    0.0884414811909
287 288 289

    References
    ----------
290
    .. [freeman-set-1977] Linton C. Freeman, "A Set of Measures of Centrality
291 292 293
       Based on Betweenness", Sociometry, Vol. 40, No. 1,  pp. 35-41 (1977)
    """

Tiago Peixoto's avatar
Tiago Peixoto committed
294
    return libgraph_tool_centrality.\
295
           get_central_point_dominance(g._Graph__graph,
Tiago Peixoto's avatar
Tiago Peixoto committed
296 297
                                       _prop("v", g, betweenness))

298 299

def eigentrust(g, trust_map, vprop=None, norm=False, epslon=1e-6, max_iter=0,
Tiago Peixoto's avatar
Tiago Peixoto committed
300
               ret_iter=False):
301 302 303 304 305
    r"""
    Calculate the eigentrust centrality of each vertex in the graph.

    Parameters
    ----------
306
    g : :class:`~graph_tool.Graph`
307
        Graph to be used.
308
    trust_map : :class:`~graph_tool.PropertyMap`
309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324
        Edge property map with the values of trust associated with each
        edge. The values must not lie in the range [0,1].
    vprop : PropertyMap, optional (default: None)
        Vertex property map where the values of eigentrust must be stored.
    norm : bool, optional (default: false)
        Norm eigentrust values so that the total sum equals 1.
    epslon : float, optional (default: 1e-6)
        Convergence condition. The iteration will stop if the total delta of all
        vertices are below this value.
    max_iter : int, optional (default: None)
        If supplied, this will limit the total number of iterations.
    ret_iter : bool, optional (default: False)
        If true, the total number of iterations is also returned.

    Returns
    -------
325
    eigentrust : A vertex property map containing the eigentrust values.
326 327 328 329 330 331 332 333 334

    See Also
    --------
    betweenness: betweenness centrality
    pagerank: PageRank centrality
    absolute_trust: absolute trust centrality

    Notes
    -----
335
    The eigentrust [kamvar-eigentrust-2003]_ values :math:`t_i` correspond the
336 337
    following limit

338 339
    .. math::

340 341 342 343 344
        \mathbf{t} = \lim_{n\to\infty} \left(C^T\right)^n \mathbf{c}

    where :math:`c_i = 1/|V|` and the elements of the matrix :math:`C` are the
    normalized trust values:

345 346
    .. math::

347 348 349 350 351 352 353 354 355 356
        c_{ij} = \frac{\max(s_{ij},0)}{\sum_{j} \max(s_{ij}, 0)}

    The algorithm has a topology-dependent complexity.

    If enabled during compilation, this algorithm runs in parallel.

    Examples
    --------
    >>> from numpy.random import poisson, random, seed
    >>> seed(42)
357
    >>> g = gt.random_graph(100, lambda: (poisson(3), poisson(3)))
358 359
    >>> trust = g.new_edge_property("double")
    >>> trust.get_array()[:] = random(g.num_edges())*42
360
    >>> t = gt.eigentrust(g, trust, norm=True)
361
    >>> print t.get_array()
Tiago Peixoto's avatar
Tiago Peixoto committed
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
    [  5.51422638e-03   1.12397965e-02   2.34959294e-04   6.32738574e-03
       0.00000000e+00   6.34804836e-03   2.67885424e-03   4.02497751e-03
       1.67943467e-02   6.46196106e-03   1.92402451e-02   9.04032352e-04
       9.70843104e-03   1.40319816e-02   1.04995777e-02   2.86712231e-02
       2.47285894e-02   2.38394469e-02   7.06936059e-03   9.45794717e-03
       2.09970054e-05   1.64768298e-03   0.00000000e+00   1.19346706e-03
       6.88434371e-03   5.36337333e-03   2.08428677e-02   2.85813783e-03
       1.10564670e-02   3.16345060e-04   5.25737238e-03   5.43761445e-03
       7.98048389e-03   7.95939648e-03   2.23891858e-02   5.68630666e-03
       2.09300588e-02   4.28902068e-03   1.70833078e-03   2.37814042e-02
       0.00000000e+00   1.20805010e-03   1.29713483e-02   5.73021992e-03
       8.71093674e-03   7.77661067e-03   8.76489806e-04   2.38519385e-02
       3.53225723e-03   8.46948906e-03   5.09874234e-03   2.44547150e-02
       1.32342629e-02   1.80085559e-03   4.37189381e-03   1.18195253e-02
       1.62748861e-02   1.83200678e-04   1.09745025e-02   1.47544090e-03
       3.34512926e-02   1.58885132e-03   1.13128910e-03   3.04944830e-02
       4.22684975e-03   0.00000000e+00   9.89654274e-04   4.25927156e-03
       2.34516214e-02   4.91370905e-03   2.29366664e-02   0.00000000e+00
       6.83407601e-03   1.60508753e-02   1.62762068e-03   3.94324856e-02
       2.84109571e-02   8.81167727e-04   2.16999908e-02   1.28688125e-02
       1.10825963e-02   2.64915564e-03   2.88711928e-03   0.00000000e+00
       4.24392252e-03   9.38398819e-03   0.00000000e+00   1.74508371e-02
       3.26594153e-02   4.07188867e-02   3.20678152e-03   6.35046287e-03
       8.07061556e-03   5.08505374e-03   3.27300367e-03   3.30989070e-03
       2.30651195e-02   4.20338525e-03   5.04332662e-03   3.58731532e-02]
387 388 389

    References
    ----------
390
    .. [kamvar-eigentrust-2003] S. D. Kamvar, M. T. Schlosser, H. Garcia-Molina
391 392 393 394 395
       "The eigentrust algorithm for reputation management in p2p networks",
       Proceedings of the 12th international conference on World Wide Web,
       Pages: 640 - 651, 2003
    """

Tiago Peixoto's avatar
Tiago Peixoto committed
396 397
    if vprop == None:
        vprop = g.new_vertex_property("double")
398 399 400 401 402 403 404 405 406 407 408
    i = libgraph_tool_centrality.\
           get_eigentrust(g._Graph__graph, _prop("e", g, trust_map),
                          _prop("v", g, vprop), epslon, max_iter)
    if norm:
        vprop.get_array()[:] /= sum(vprop.get_array())

    if ret_iter:
        return vprop, i
    else:
        return vprop

409
def absolute_trust(g, trust_map, source, target = None, vprop=None):
410
    r"""
411 412
    Calculate the absolute trust centrality of each vertex in the graph, from a
    given source.
413 414 415

    Parameters
    ----------
416
    g : :class:`~graph_tool.Graph`
417
        Graph to be used.
418
    trust_map : :class:`~graph_tool.PropertyMap`
419 420
        Edge property map with the values of trust associated with each
        edge. The values must lie in the range [0,1].
421 422 423 424 425 426
    source : Vertex
        A vertex which is used the as the source for gathering trust values.
    target : Vertex (optional, default: None)
        A vertex which is used the as the only target for which the trust value
        will be calculated. If left unspecified, the trust values for all
        targets are computed.
427
    vprop : :class:`~graph_tool.PropertyMap`, optional (default: None)
428
        A vertex property map where the values of trust for each source
429
        must be stored.
430 431 432

    Returns
    -------
433
    absolute_trust : :class:`~graph_tool.PropertyMap` or float
434
        A vertex property map containing the absolute trust vector from the
435 436 437
        source vertex to the rest of the network. If `target` is specified, the
        result is a single float, with the corresponding trust value for the
        target.
438

439 440 441 442 443 444 445 446 447 448
    See Also
    --------
    eigentrust: eigentrust centrality
    betweenness: betweenness centrality
    pagerank: PageRank centrality

    Notes
    -----
    The absolute trust between vertices i and j is defined as

449 450
    .. math::

451 452
        t_{ij} = \frac{\sum_m A_{m,j} w^2_{G\setminus\{j\}}(i\to m)c_{m,j}}
                 {\sum_m A_{m,j} w_{G\setminus\{j\}}(i\to m)}
453

454 455 456
    where :math:`A_{ij}` is the adjacency matrix, :math:`c_{ij}` is the direct
    trust from i to j, and :math:`w_G(i\to j)` is the weight of the path with
    maximum weight from i to j, computed as
Tiago Peixoto's avatar
Tiago Peixoto committed
457

458 459
    .. math::

460
       w_G(i\to j) = \prod_{e\in i\to j} c_e.
461

462 463 464 465 466 467
    The algorithm measures the absolute trust by finding the paths with maximum
    weight, using Dijkstra's algorithm, to all in-neighbours of a given
    target. This search needs to be performed repeatedly for every target, since
    it needs to be removed from the graph first. The resulting complexity is
    therefore :math:`O(N^2\log N)` for all targets, and :math:`O(N\log N)` for a
    single target.
468 469 470 471 472 473 474

    If enabled during compilation, this algorithm runs in parallel.

    Examples
    --------
    >>> from numpy.random import poisson, random, seed
    >>> seed(42)
475
    >>> g = gt.random_graph(100, lambda: (poisson(3), poisson(3)))
476
    >>> trust = g.new_edge_property("double")
477
    >>> trust.a = random(g.num_edges())
478 479
    >>> t = gt.absolute_trust(g, trust, source=g.vertex(0))
    >>> print t.a
Tiago Peixoto's avatar
Tiago Peixoto committed
480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496
    [ 0.16260667  0.04129912  0.13735376  0.19146125  0.          0.09147461
      0.10371912  0.12465511  0.24631221  0.0603916   0.2375385   0.06637879
      0.08897662  0.0800988   0.05250601  0.66759022  0.09368793  0.08275437
      0.13674709  0.15553915  0.01376162  0.417068    0.          0.06096886
      0.08746817  0.39380693  0.09215297  0.09575144  0.15594162  0.04008874
      0.05483972  0.05691086  0.13571077  0.32376012  0.22477937  0.06347962
      0.10445085  0.19447845  0.38007043  0.13810585  0.          0.08451096
      0.06648153  0.18479174  0.13003649  0.14850631  0.00320603  0.1074644
      0.12088162  0.06792678  0.08472666  0.2002143   0.25963204  0.37838425
      0.03089371  0.18389694  0.39420339  0.03348093  0.11483196  0.0656204
      0.14206403  0.07066434  0.25168986  0.07040126  0.04870569  0.
      0.09861349  0.03882069  0.1105267   0.07951823  0.08748441  0.
      0.08393443  0.11121719  0.21903223  0.25529628  0.0414386   0.03695558
      0.17664854  0.05143033  0.11735779  0.06525968  0.19600919  0.          0.1220922
      0.33330041  0.          0.28595961  0.14526678  0.12514885  0.089524
      0.40738962  0.03719195  0.54409979  0.06247424  0.10660136  0.11674385
      0.13218144  0.02214988  0.23215937]
497
    """
Tiago Peixoto's avatar
Tiago Peixoto committed
498 499

    if vprop == None:
500
        vprop = g.new_vertex_property("double")
501

502
    source = g.vertex_index[source]
503

504 505 506 507
    if target == None:
        target = -1
    else:
        target = g.vertex_index[target]
508

509 510 511 512 513
    libgraph_tool_centrality.\
            get_absolute_trust(g._Graph__graph, source, target,
                               _prop("e", g, trust_map), _prop("v", g, vprop))
    if target != -1:
        return vprop.a[target]
514
    return vprop
Tiago Peixoto's avatar
Tiago Peixoto committed
515