Yikes, something goes wrong in col.rs@test_col.

This only happens for linux-conda-test, I believe that this happens because conda uses the newest SCIP (10.0), whilst linux-bundled-test uses v9.2.4.

As you guys are probably pretty busy with the new release of SCIP, I will try and check later this week if I can compile locally against SCIP 10.0!

Yikes, something goes wrong in col.rs@test_col.

This only happens for linux-conda-test, I believe that this happens because conda uses the newest SCIP (10.0), whilst linux-bundled-test uses v9.2.4.

As you guys are probably pretty busy with the new release of SCIP, I will try and check later this week if I can compile locally against SCIP 10.0!

Feel free to change the pipeline for now to use SCIP 9.2.4.

Thank you for your contribution! It looks good to me on a first glance, I'll take another look sometime in the next days.

Maybe unrelated, feel free to not answer; I tried creating a custom branching-rule that should work together with the already existing ones in SCIP. All the branching decisions that my rule makes throughout the tree, should be passed on to all of the node's descendants. For this, I keep a BTreeMap that is indexed by the node index and contains all the decisions made for a node (similar to the Ryan-Foster branching from the examples).

However, if SCIP uses its own branching rules, then the indexes of the Map were not updated. Therefore I added an event handler that is called after branching. This way I can simply update the Map manually. Do you think that this is the best way of tackling this issue? Or should I use the SCIP_BranchruleData that is written about in the documentation?

I can't think of an easier solution than what you did. As far as I know, the SCIP_BranchruleData is a placeholder type that each plugin overrides with its own type, and this is what russcip uses to give ownership of the type that implements the BranchRule trait.

I can't think of an easier solution than what you did. As far as I know, the SCIP_BranchruleData is a placeholder type that each plugin overrides with its own type, and this is what russcip uses to give ownership of the type that implements the BranchRule trait.

Alright! Good to know, thanks!

Codecov Report

❌ Patch coverage is 92.30769% with 2 lines in your changes missing coverage. Please review.
✅ Project coverage is 76.49%. Comparing base (10c29b4) to head (844223e).
⚠️ Report is 6 commits behind head on main.

@@            Coverage Diff             @@
##             main     #268      +/-   ##
==========================================
+ Coverage   76.37%   76.49%   +0.12%     
==========================================
  Files          29       29              
  Lines        3830     3871      +41     
==========================================
+ Hits         2925     2961      +36     
- Misses        905      910       +5     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.