Provides an implicit conversion to a DiameterApproximator object
that allows to call the diameterApprox
function on a Graph instance
Provides an implicit conversion to a DiameterApproximator object
that allows to call the diameterApprox
function on a Graph instance
any vertex data is ignored
the graph to wrap in a DiameterApproximator
a DiameterApproximator wrapping the given graph
Runs the diameter approximation algorithm.
Runs the diameter approximation algorithm. The two parameters have the following meaning:
- target
: this is the size of the quotient graph that will be built
by the underlying clustering algorithm. It depends on the size of
the local memory of the machines. The last step of the algorithm
computes the diameter of a graph of size target
. Higher values of
target
can result is shorter running times, whereas smaller ones
require less memory.
- delta
: this parameter, representing a distance,
controls the number of nodes and edges that can be active in each
step of the algorithm. Intuitively, higher values will result in
fewer but slower rounds; smaller values will perform more shorter
rounds. In any case, this parameter is taken as a hint by the
algorithm, that then auto-tunes itself.
Further details are provided in the companion papers.
data associated to vertices is ignored
a weighted graph
the target size for the quotient graph
the parameter for the underlying delta-stepping like algorithm
an approximation to the diameter of the graph
Runs the algorithm with the specified delta parameter.
Runs the algorithm with the specified delta parameter.
The algorithm runs with the default value for target
, 4000.
data associated to vertices is ignored
a weighted graph
the parameter for the underlying delta-stepping like algorithm
an approximation to the diameter of the graph
Runs the clustering algorithm with the specified target quotient size.
Runs the clustering algorithm with the specified target quotient size.
In this case delta
defaults to the average edge weight.
data associated to vertices is ignored
a weighted graph
the target size for the quotient graph
an approximation to the diameter of the graph
Runs the approximation algorithm with default values.
Runs the approximation algorithm with default values.
The default values are 4000
for target
and the average edge
weight for delta
.
data associated to vertices is ignored
a weighted graph
an approximation to the diameter of the graph
Functions to approximate the diameter of large graphs. The algorithm implemented here is based on the ones described in the papers
by Matteo Ceccarello, Andrea Pietracaprina, Geppino Pucci, and Eli Upfal
by Matteo Ceccarello, Andrea Pietracaprina, Geppino Pucci, and Eli Upfal
The functions provided by this object work on weighted graphs where edge weights are of type Distance (i.e.
Double
). The data associated with vertices is ignored.