Solver Parameter - Committed Student Conflicts

[David Berndt]

Hello.  Can you explain the different between Load, Compute and Ignore as a solver parameter for Committed Student Conflicts.  How does this choice impact how the solver runs and attempts to place students, both when using curricula and student plans to evaluate student demands?

In my current approach I have most of the classes committed and am loading one or two solver groups of courses (1 or 2 departments) to make changes and improvements in subsets of the larger schedule.  I'd like the solver to account for other committed assignments, but attempt to move students around between committed and uncommitted courses to help find the best result.  How does my choice here limit what the solver can do with a subset of classes are chosen.  Thanks!

[Tomáš Müller]

Hi David,

This parameter is only used if you have multiple solver groups, and some of them already have a committed solution. Imagine that a student requests two courses, a CS course (timetabled by the CS department) and a MATH course (timetabled by the MATH department). Now, let's say that the MATH course is already timetabled, and you are loading CS into the solver. If the parameter is set to Ignore, the other course will get ignored (there will not be any student conflicts between the two courses reported/minimized). If the parameter is set to Load, the solver will load the enrollments as they were in the other problem (MATH) -- so, if the student was placed in Lec 1 of the MATH course, there could be committed student conflicts for the classes that the student is in, if they overlap with the Lec 1 section of the MATH course. This, however, only works for student course demands that work with actual students (all but curricula), if you use Curricula Course Demands, there are no student enrollments saved as the solver is working with students that are made up following the curricula during the load. So, if you use Curricula Course Demands, the solver will fall back to Compute even when Load was selected. The last option, Compute, will also have the student take both courses. However, the solver will load the demands but may assign the students to different sections of the courses of the other (committed) problems, as it computes the individual student-class assignments rather than loading them from the solution(s).

Best regards,

Tomas

On Fri, Mar 14, 2025 at 6:11 PM David Berndt <dber...@gmail.com> wrote:

Hello.  Can you explain the different between Load, Compute and Ignore as a solver parameter for Committed Student Conflicts.  How does this choice impact how the solver runs and attempts to place students, both when using curricula and student plans to evaluate student demands?

In my current approach I have most of the classes committed and am loading one or two solver groups of courses (1 or 2 departments) to make changes and improvements in subsets of the larger schedule.  I'd like the solver to account for other committed assignments, but attempt to move students around between committed and uncommitted courses to help find the best result.  How does my choice here limit what the solver can do with a subset of classes are chosen.  Thanks!

--
You received this message because you are subscribed to the Google Groups "UniTime Discussion Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to unitime-discu...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/unitime-discuss/e2896e86-a7a2-4c07-af0f-8955dd8ed128n%40googlegroups.com.

[David Berndt]

Very interesting!  Yes the scenario you outline if in line with what I'm attempting.  It seems like Compute is the setting I should use as I'd want the solver to try to work in the best temporary schedule placement for the students between committed courses those those currently running through the solver.   

By the way, I'm attending AACRAO in Seattle coming up.  If you or any other Unitime users you know of are in attendance I'd love to meet up.  Feel free to reach out to me.  Thanks!