Timestamp:
Nov 11, 2013, 1:26:06 PM (12 years ago)
Author:
Frederik Heber <heber@…>
Children:
15c7f52
Parents:
d0ab01
git-author:
Frederik Heber <heber@…> (11/11/13 09:49:19)
git-committer:
Frederik Heber <heber@…> (11/11/13 13:26:06)
Message:

FIX: CyclicStructureAnalysis falsely used DFS, skipped some cycles.

  • FIX: CyclicStructureAnalysis did use DFS instead of BFS for finding cycles. Note that CyclicStructureAnalysis with coronene would find supercycles and not the smaller interconnected ones.
  • FIX: Cycles were skipped when all bonds marked cyclic, not enough. In interconnected aromatic rings, bonds may very well be marked as cyclic from earlier extraction of cycles and yet the specific cycle might not have been found yet (e.g. coronene). In this case we now check whether this particular cycle has already been extracted and only skip if so.
  • TESTFIX: added new fragmentation regression tests on some metallic systems.
  • this is mainly for regression on bond graph detection and cycle analysis.
(No files)

Note: See TracChangeset for help on using the changeset viewer.