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
Describe the bug
A double free happens within the C++ solver, when a factor is given that contains a variable that is marked as optimized but never appears in the actual problem statement.
To Reproduce
Use any optimization problem formulation in python using the Factor class.
Remove an arbitrary symbolic variable from your problem entirely, but leave its definition.
Add this symbolic variable to the optimized_keys in factor.generate()
Use the generate code in the LM solver.
Watch it crash.
Expected behavior
Optimized variable should simply be kept at its initial value,
Environment (please complete the following information):
Skydio developer station
The text was updated successfully, but these errors were encountered:
Previously it was possible to construct an optimization problem with
sparse factors with symbolic zeros on the diagonal. We now explicitly
add entries to the symbolic diagonal so that the sparsity pattern is
correct after damping
Fixessymforce-org#380
Topic: sf-sparse-zeroes
Reviewers: ryan-b,chao,nathan,brad,michael-f,philipp
GitOrigin-RevId: 6d4b44cbf6fcd977515e557c9a95eab6a5ca1b6a
Describe the bug
A double free happens within the C++ solver, when a factor is given that contains a variable that is marked as optimized but never appears in the actual problem statement.
To Reproduce
Factor
class.optimized_keys
infactor.generate()
Expected behavior
Optimized variable should simply be kept at its initial value,
Environment (please complete the following information):
The text was updated successfully, but these errors were encountered: