You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
dehann
changed the title
Should multiple solvers use the same CSMC and Tree?
Should multiple simultaneous solvers use the same CSMC and Tree?
Apr 4, 2020
dehann
changed the title
Should multiple simultaneous solvers use the same CSMC and Tree?
Should simultaneous solvers use the same CSMC and Tree?
Apr 4, 2020
dehann
changed the title
Should simultaneous solvers use the same CSMC and Tree?
Should simultaneous solvers use the same CSMC / Tree?
Apr 4, 2020
dehann
changed the title
Should simultaneous solvers use the same CSMC / Tree?
Should simultaneous solvers use same CSMC / Tree?
Apr 4, 2020
dehann
changed the title
Should simultaneous solvers use same CSMC / Tree?
Simultaneous solvers use same CSMC / Tree?
Apr 4, 2020
Note, building a tree requires temporary changes in the factor graph object and hence only one tree can be built at a time, however, multiple trees can be used simultaneously on the same factor graph object after they have been built.
For most flexibility it might be best to try support both models as far possible (short of rebuilding the tree building process in the short term).
For trees, work towards a DFG style where TreeClique has all data common all solvers, but then TreeCliq contains a field as ::Dict{Symbol, BTND} used for each of the different solvers. This implies elements like potentials (#680) should move up to TreeClique, etc.
Picking up from #459 (comment)
The text was updated successfully, but these errors were encountered: