Branching in HINT_SEARCH with incomplete hint.

20 views
Skip to first unread message

Samrat Kadge

unread,
May 10, 2026, 12:13:35 PM (2 days ago) May 10
to or-tools-discuss
Provided an incomplete hint,

1) Will the search tree be started with variables listed in the hint,
similar to PARTIAL_FIXED_SEARCH ?

2) Apart from using larger hint_conflict_limit,
are there are any tips to complete the hint (and quickly)

Context:
I solve a rostering problem using many minizinc globals
e.g global_cardinality, cost_regular, sliding_sum

roster  is my independent variable, all other variables being dependent.

Log details
%% The solution hint is incomplete: 29070 out of 439570 non fixed variables hinted.
%% The solution hint is incomplete: 4152 out of 1908198 non fixed variables hinted.

Thanks,
Samrat

Laurent Perron

unread,
May 10, 2026, 12:22:55 PM (2 days ago) May 10
to or-tools...@googlegroups.com
If there is an incomplete hint, the search starts with hint_search, then after the hint conflict limit switches to the normal search.

You can use the parameter: fix_variables_to_their_hinted_value:true 
Laurent Perron | Operations Research | lpe...@google.com | (33) 1 42 68 53 00



--
You received this message because you are subscribed to the Google Groups "or-tools-discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to or-tools-discu...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/or-tools-discuss/5c7a0c9c-3b5e-48b1-8cce-dc3f7d84d3b4n%40googlegroups.com.

Samrat Kadge

unread,
May 10, 2026, 11:13:03 PM (2 days ago) May 10
to or-tools-discuss
1. I am performing lexicographic optimization, and the solution hint is from the previous stage
(whatever variables are available at the minizinc level)

2. Using fix_variables_to_their_hinted_value:true
%% Fixing 46029 variables to their value in the solution hints.
%% The solution hint is incomplete: 26076 out of 166972 non fixed variables hinted.
%% The solution hint is complete and is feasible. Its objective value is 12973.
(under a minute)

3. I want to start with this solution and find better solutions

4. Is there a similar parameter for a subsolver?

The situation gets worse with every stage of optimization. (upto an hour before a first solution)

Thanks
Reply all
Reply to author
Forward
0 new messages