Commit Graph

51 Commits

Author SHA1 Message Date
Nadine Schneider
624dd72ad6 new canon: fix in special symmetry invariant 2015-11-06 18:05:04 +01:00
Nadine Schneider
0cf0dd37ce Bugfix in SmilesWrite and some additional tests for getMolFrags function 2015-04-16 10:53:20 +02:00
Nadine Schneider
0b8b6a3690 Some performance fixes for the new canonicalization
- handling molecules/proteins with multiple fragments
- add criteria for using special symmetry invariant
2015-04-15 18:42:29 +02:00
Nadine Schneider
0132cc72e9 Merge branch 'newCanon' into master3
Conflicts:
	Code/GraphMol/SmilesParse/test.cpp
2015-04-10 10:25:47 +02:00
Nadine Schneider
b4c230c7d0 new_canon: move hanoiSort to a seperate file 2015-04-10 09:47:17 +02:00
Nadine Schneider
ca4bcca6ae Include atom-mapping numbers if present in the canonicalization 2015-04-10 09:47:16 +02:00
Nadine Schneider
69ceace885 Bugfix in SpecialChiralityAtomCompareFunctor and additional tests
Conflicts:

	Code/GraphMol/hanoitest.cpp
2015-04-10 09:47:09 +02:00
Nadine Schneider
7e59908121 SmilesWrite: introduce new function for fragments 2015-03-20 16:38:23 +01:00
Nadine Schneider
e013399bb9 Bugfixes and new invariant for chiralAtoms 2015-03-06 15:47:08 +01:00
Nadine Schneider
0d90e31c46 New canon: new atom invariant also including neighbors 2015-02-12 18:08:03 +01:00
Nadine Schneider
465c7cb4c6 new_canon: Delete numRingMember invariant 2015-02-12 18:06:27 +01:00
Nadine Schneider
66fe6cf10f update and new canon tests 2015-02-12 13:05:59 +01:00
Nadine Schneider
d82f30f5d4 Optimization canon 2015-02-02 12:53:19 +01:00
Greg Landrum
38cab28eb6 add a couple fullerenes as renumbering tests 2015-01-28 01:26:15 +01:00
Greg Landrum
002d6f7cb0 minor optimization 2015-01-16 14:26:22 +01:00
Greg Landrum
27c6c61d9b At this point the hanoiTests pass and all 140K ZNP compounds pass the 50x torture test. 2015-01-16 05:08:36 +01:00
Greg Landrum
baf26c053c not fixed; still lots of debugging printing; backup commit 2015-01-09 06:33:49 +01:00
Greg Landrum
1f4c2e915c fix a nasty canonicalization problem:
need to be sure to sort neighbors by their ranks in a *decreasing* order
2015-01-07 20:46:08 +01:00
Greg Landrum
756404a0b0 re-enable some tests 2015-01-05 09:25:51 +01:00
Greg Landrum
48c5d0d6ff At this point all C++ tests pass 2015-01-03 06:42:54 +00:00
Greg Landrum
5a3e65fd30 at least the basics of stereochemistry are working
all hanoitest.cpp tests passing
chiralityTest.cpp failes still, so we are not done yet
2014-12-23 08:05:47 +00:00
Greg Landrum
ab0f1dcf60 rename to new_canon (probably should come up with a better name, but this will do for now) 2014-12-22 07:56:26 +00:00
Greg Landrum
cad05ca184 get rid of RankAtoms.cpp and
move the remaining functionality from RankAtoms.h into the new file
RDGeneral/Ranking.h
2014-12-22 07:35:58 +00:00
Greg Landrum
d11deb813b all hanoitest.cpp tests pass at this point.
The challenging dummy-atom case is commented out.
2014-12-22 06:47:15 +00:00
Greg Landrum
b286d4e601 better error reporting; but still not working 2014-08-27 04:05:45 +02:00
Greg Landrum
2f701327f2 backup 2014-08-25 05:50:50 +02:00
Greg Landrum
0d21cdc7c9 backup commit 2014-08-24 16:24:31 +02:00
Greg Landrum
643b927c25 backup before messing with chirality 2014-08-24 13:18:58 +02:00
Greg Landrum
329bf48b77 fix a pathological symmetry case by modifying the atom invariants 2014-08-24 10:51:35 +02:00
Greg Landrum
88ce84aa9c certainly should not be including the chiral tag itself in the comparison 2014-08-23 09:46:35 +02:00
Greg Landrum
4a4d724466 hanoi tests passing again 2014-08-23 03:22:09 +02:00
Greg Landrum
f9906e055b add in a few more tests with ring stereo. These currently fail. 2014-08-23 02:02:51 +02:00
Greg Landrum
3a9169d642 bug fix from Roger 2014-08-23 02:01:28 +02:00
Greg Landrum
101a5879d7 focus the tests being run 2014-08-08 09:16:01 +02:00
Greg Landrum
ee66d4e068 backup commit 2014-08-08 06:44:56 +02:00
Greg Landrum
e170195421 backup commit; this does not completely work yet 2014-08-03 08:01:58 +02:00
Greg Landrum
787d9d4c89 some less successful experiments with the node changed optimization added, but commented out 2014-07-31 04:34:25 +02:00
Greg Landrum
3edfafec87 about ready to add and tests the node-changed optimization 2014-07-30 03:34:39 +02:00
Greg Landrum
9906fd7842 get ready for the node changed optimization 2014-07-29 04:48:22 +02:00
Greg Landrum
a01f3fc042 rename to rankMolAtoms; add breakTies option 2014-07-27 07:44:34 +02:00
Greg Landrum
4d8041ead5 everything works except the renumbering test 2014-07-26 16:36:30 +02:00
Greg Landrum
5c3369805f makes it through all tests so far 2014-07-26 15:56:38 +02:00
Greg Landrum
faf805f9cc get cleverer about atom comparison 2014-07-26 08:38:23 +02:00
Greg Landrum
1bd244c28e steps towards having atoms compare themselves 2014-07-26 04:11:03 +02:00
Greg Landrum
2b74d8c78f steps towards having atoms compare themselves 2014-07-26 04:04:40 +02:00
Greg Landrum
c78effb0d4 fix a misconception that was leading to the test1() failures 2014-07-26 03:40:25 +02:00
Greg Landrum
c4b46e6677 basic functionality done; basic testing done; time for larger-scale tests and performance comparison 2014-07-25 05:52:10 +02:00
Greg Landrum
db06606454 first pass at tie-breaking is working 2014-07-25 05:15:34 +02:00
Greg Landrum
7b61ea6cd5 another big step: now we have uniquification propagating through molecules.
Note that the Morgan-encoding step is definitely not optimal
2014-07-25 05:02:51 +02:00
Greg Landrum
d3ab885b8a the random int vector test no longer passes, but we now have a lot of molecule stuff working (basic class assignment at least 2014-07-24 14:17:04 +02:00