Diffusion MOLEKUEHL (master) History

R11301/onepass.py
master1c17dba2d7641ada80e17d7…

Jan 6 2022

Changed preprocess.py for old qm7 data to have same structure as the new ones.
1c17dba2d764
·matt on Jan 6 2022, 23:05.

Jan 4 2022

fix CM
9d7adea85e44
·puckvg on Jan 4 2022, 16:31.

Dec 21 2021

aCM doesnt work
16d8a316e929
·puckvg on Dec 21 2021, 17:54.
Added modifiable constant for fragmentation penalty
d00e36abbefe
·matt on Dec 21 2021, 15:03.
Support for 3 new vector representations. Model parameters to avoid getting…
fa27a77d2a83
·matt on Dec 21 2021, 14:20.

Dec 20 2021

fix merge
8e27e829e380
·puckvg on Dec 20 2021, 15:49.
update database required change to aqml code
d882c03886a3
·puckvg on Dec 20 2021, 15:48.
Force filling solution pool with best solutions.
a3515913792d
·matt on Dec 20 2021, 15:27.

Dec 15 2021

Added SLATM representation support
bcbc7a7b5891
·matt on Dec 15 2021, 21:12.

Dec 14 2021

changed database sizes to 7/8/9 max heavy atom fragments respectively
5cc7e9b3d9f7
·puckvg on Dec 14 2021, 17:16.

Dec 6 2021

Added partial molecule penalty
1a514eeaa66c
·matt on Dec 6 2021, 15:55.

Dec 5 2021

fixed objective value
a6c6e4a68e80
·matt on Dec 5 2021, 01:43.

Dec 3 2021

Added back molecule duplication - limited to 3 copies for perfomance
3005d8c71b0b
·matt on Dec 3 2021, 14:57.

Nov 30 2021

Switched to ideal databases. Removed duplicate molecules until objective…
2ae2563a46c3
·matt on Nov 30 2021, 01:36.

Nov 29 2021

fixing merge issue
327fab3e8a81
·puckvg on Nov 29 2021, 11:13.
add optimal databases for each target
83ee1e954e70
·puckvg on Nov 29 2021, 11:12.

Nov 25 2021

Removed (commented) the connectivity conditions that were based on adjacency…
3744c8b0cf63
·matt on Nov 25 2021, 16:38.

Nov 23 2021

fix sign of constraint in obj
361d72e931d5
·puckvg on Nov 23 2021, 13:57.
fix merge
74a6c6e3bef1
·puckvg on Nov 23 2021, 13:56.
update
2da90f129e25
·puckvg on Nov 23 2021, 13:49.

Nov 22 2021

Added preprocessing of the connectivity matrices to improve on runtime; various…
b1b6ccce0c15
·matt on Nov 22 2021, 15:25.

Nov 20 2021

Memory improvements; readded taking molecules multiple times.
23da2886dc0f
·matt on Nov 20 2021, 18:49.

Nov 19 2021

Added weak connectivity conditions on fragments of molecule taken. Issues…
3ae7e7d2c4ca
·matt on Nov 19 2021, 19:41.

Nov 16 2021

Removed partial target matching; added parameter for size of molecule fragment
3d34d30e4e22
·matt on Nov 16 2021, 13:44.

Nov 12 2021

Improved output runtime; now prints old positions (indices+1) before…
1463d851f136
·matt on Nov 12 2021, 13:37.

Nov 11 2021

Modifed preprocessing to remove hydrogens; various changes on onepass (i forgot…
cad06c9aec83
·matt on Nov 11 2021, 16:28.

Nov 10 2021

Merge branch 'master' of https://c4science.ch/diffusion/11301/molekuehl
bd45bf0be154
·matt on Nov 10 2021, 14:48.
Enhanced output; fixed size constraints.
51ebee2f6454
·matt on Nov 10 2021, 14:42.
adding xyz files for targets and database to look
aacc423084e7
·puckvg on Nov 10 2021, 12:08.

Nov 9 2021

Allow partial molecule with threshold; force portion of target to be assigned…
6b022f2e9bb6
·matt on Nov 9 2021, 16:22.
First sketch of preprocessing; allowed taking molecules multiple times
ee3c54d0dd06
·matt on Nov 9 2021, 14:45.

Nov 8 2021

Added output for stored solutions; time and gap limits.
a440a7a3f8b0
·matt on Nov 8 2021, 16:51.

Nov 5 2021

Few changes to improve running time
d457ff7a09fe
·matt on Nov 5 2021, 19:19.

Nov 4 2021

Code refactoring
d3cf1ed4215f
·matt on Nov 4 2021, 14:12.

Nov 3 2021

New file aiming for one-pass algorithm based on gurobi.py.
2870d9c1d762
·matt on Nov 3 2021, 17:43.