Closed sebhtml closed 11 years ago
If I turn off the null edge purger, the edge distribution becomes:
mpiexec -output-filename degree-bug \ -n $NSLOTS ~/Ray -p u1.fasta u2.fasta -o degree-bug \ -bloom-filter-bits 0
patch:
diff --git a/code/plugin_EdgePurger/EdgePurgerWorker.cpp b/code/plugin_EdgePurger/EdgePurgerWorker.cpp index 367edd7..d6e0eca 100644 --- a/code/plugin_EdgePurger/EdgePurgerWorker.cpp +++ b/code/plugin_EdgePurger/EdgePurgerWorker.cpp @@ -48,7 +48,7 @@ void EdgePurgerWorker::work(){ CoverageDepth coverage=response[0]; - if(coveragegetMinimumCoverageToStore()){ + if(false &&coverage getMinimumCoverageToStore()){ m_vertex->deleteIngoingEdge(&m_currentKmer,&vertex,m_parameters->getWordSize()); } m_iterator++; @@ -81,7 +81,7 @@ void EdgePurgerWorker::work(){ CoverageDepth coverage=response[0]; - if(coverage getMinimumCoverageToStore()){ + if(false &&coverage getMinimumCoverageToStore()){ m_vertex->deleteOutgoingEdge(&m_currentKmer,&vertex,m_parameters->getWordSize()); } m_iterator++;
Result:
0 0 5 0 1 6383 0 2 0 0 3 0 0 4 0 1 0 6174 1 1 363860 1 2 9155 1 3 197 1 4 2 2 0 0 2 1 9155 2 2 738 2 3 18 2 4 0 3 0 0 3 1 197 3 2 18 3 3 2 3 4 0 4 0 0 4 1 2 4 2 0 4 3 0 4 4 0
Therefore the problem is before the purging of null edges.
With fix:
0 0 0 0 1 6179 0 2 0 0 3 0 0 4 0 1 0 6179 1 1 364064 1 2 9155 1 3 197 1 4 2 2 0 0 2 1 9155 2 2 738 2 3 18 2 4 0 3 0 0 3 1 197 3 2 18 3 3 2 3 4 0 4 0 0 4 1 2 4 2 0 4 3 0 4 4 0
dda2d53775b63582eb0e9907b5b8fd5841f1a5b4
Setup:
genomeX.fasta / 74 kb
16 k sequences, 1% error
Coverage distribution:
2 246698 3 6086 4 7978 5 14022 6 19264 7 21844 8 21526 9 18366 10 15104 11 10252 12 6788 13 3956 14 2158 15 1034 16 518 17 190 18 58 19 30 20 28 21 6
degree distribution
(0,0) should be even
(0,1) and (1,0) should be the same
0 0 225649 0 1 10869 0 2 0 0 3 0 0 4 0 1 0 10862 1 1 148206 1 2 159 1 3 0 1 4 0 2 0 0 2 1 159 2 2 2 2 3 0 2 4 0 3 0 0 3 1 0 3 2 0 3 3 0 3 4 0 4 0 0 4 1 0 4 2 0 4 3 0 4 4 0
This bug can cause this:
Error: The seed contains a choice not supported by the graph.
Because any edge should have its reverse complement edge too