pylogeny.heuristic module

Define the interface for a heuristic in order to implement any manner of heuristic for a combinatorial problem that can be abstracted into a state graph. In this case, a phylogenetic tree space.

class pylogeny.heuristic.RAxMLIdentify(ls, startNode, workdir='.rxml')[source]

Bases: pylogeny.heuristic.phylogeneticLinearHeuristic

RAxML-driven landscape evaluation of intermediate checkpoint trees output from the RAxML executable.

__init__(ls, startNode, workdir='.rxml')[source]

Initialize this heuristic.

Parameters:
  • ls (a landscape.landscape object) – a landscape object
  • startNode (a node (dictionary) from the landscape (getNode())) – what node to start with
explore()[source]

Explore using RAxML.

Returns:None; landscape is modified.
class pylogeny.heuristic.heuristic(G=None, start=None)[source]

Bases: object

A base interface for a heuristic that explores a state graph.

__init__(G=None, start=None)[source]
explore()[source]
getStartState()[source]
getStateGraph()[source]
class pylogeny.heuristic.likelihoodGreedy(ls, startNode)[source]

Bases: pylogeny.heuristic.phylogeneticLinearHeuristic

Greedy (hill-climbing) landscape exploration by comparsion of likelihood.

__init__(ls, startNode)[source]

Initialize this heuristic.

Parameters:
  • ls (a landscape.landscape object) – a landscape object
  • startNode (a node (dictionary) from the landscape (getNode())) – what node to start with
explore()[source]

Perform greedy search of the landscape using a method of greed via likelihood.

Returns:None; landscape is modified
class pylogeny.heuristic.parsimonyGreedy(ls, startNode)[source]

Bases: pylogeny.heuristic.phylogeneticLinearHeuristic

Greedy (hill-climbing) landscape exploration by comparsion of parsimony.

__init__(ls, startNode)[source]

Initialize this heuristic.

Parameters:
  • ls (a landscape.landscape object) – a landscape object
  • startNode (a node (dictionary) from the landscape (getNode())) – what node to start with
explore()[source]

Perform greedy search of the landscape using a method of greed via parsimonious criterion.

Returns:None; landscape is modified
class pylogeny.heuristic.phylogeneticLinearHeuristic(ls, startNode)[source]

Bases: pylogeny.heuristic.heuristic

A base class for a heuristic that works on a phylogenetic landscape and only possesses a single path (of search).

__init__(ls, startNode)[source]
bestTree = None
getBestTree()[source]
getPath()[source]
path = []
class pylogeny.heuristic.smoothGreedy(ls, startNode)[source]

Bases: pylogeny.heuristic.phylogeneticLinearHeuristic

Parsimony-driven greedy landscape exploration by comparsion of likelihoods.

__init__(ls, startNode)[source]

Initialize this heuristic.

Parameters:
  • ls (a landscape.landscape object) – a landscape object
  • startNode (a node (dictionary) from the landscape (getNode())) – what node to start with
explore()[source]

Perform greedy search of the landscape using a method of greed via parsimonious criterion and then performing final smoothing via likelihood on top 10% of 1-SPR neighbors ranked on basis of parsimony.

Returns:None; landscape is modified

Previous topic

pylogeny.executable module

Next topic

pylogeny.landscape module

This Page