Ignore:
Timestamp:
Jul 8, 2015, 2:59:03 PM (4 years ago)
Author:
nmedfort
Message:

Partial roll back of Trie structure. Seemed to introduce the potential of generating a cycle in the AST. More analysis is needed here before mixing together multiple Advance DAG traversals.

File:
1 edited

Legend:

Unmodified
Added
Removed
  • icGREP/icgrep-devel/icgrep/pablo/optimizers/pablo_automultiplexing.hpp

    r4648 r4650  
    4444    DdNode * characterize(Advance * adv, DdNode * input);
    4545    bool notTransitivelyDependant(const ConstraintVertex i, const ConstraintVertex j) const;
    46     bool generateMultiplexSets(RNG & rng, unsigned k = 1);
    47     void addCandidateSet(const VertexVector & M, Trie & T, VertexVector & roots) const;
    48     void writeCandidateSet(const Trie::vertex_descriptor t, const Trie & T, VertexVector & S);
     46    bool generateCandidateSets(RNG & rng);
     47    void addCandidateSet(const VertexVector & S);
    4948    void selectMultiplexSets(RNG &);
    5049    void applySubsetConstraints();
Note: See TracChangeset for help on using the changeset viewer.