Group Visits
Set and verify how often people are assigned to selected groups.
Add a Group Visits preference
Open Scenario Setup → Preferences → Group Visits, then select Add Group Visits.
- Choose the People, Groups, and Sessions.
- Choose Group counting.
- Set the minimum, maximum, or both. Distinct counting labels these Minimum groups and Maximum groups.
- Set Penalty function and Penalty weight.
- Select Save.
Group Visits is a preference. A result may miss the target and remain valid; each miss contributes weighted cost.
Set the scope
For People, Groups, and Sessions, All includes current and future items. Selecting items creates a fixed list that does not grow when the scenario changes.
A person can accumulate visits only in selected sessions they attend. Fixed placements and other requirements still apply. For a station rotation, leave stationary facilitators outside the rotating People scope unless their own station assignments should be counted.
Choose how groups are counted
With 4 people and 3 groups, separate counting creates 12 measurements. Combined and distinct counting each create 4. For a person assigned to A, A, B, C, combined counting gives 4 visits and distinct counting gives 3 different groups. A combined count increases by at most 1 per session because a person has one assignment.
Set the visit target
Use the inclusive bounds directly:
- set only Minimum visits for an at-least target;
- set only Maximum visits for an at-most target;
- set both to different values for a range;
- set both to the same value for an exact visit count.
An empty side means that bound is not applied. Counts are non-negative whole numbers. For visit counting, the form limits the minimum to the number of selected sessions. For distinct counting, the reachable count is at most the smaller of the number of selected groups and the person’s attended, selected sessions. A larger distinct target is allowed and incurs a shortfall penalty; it is never silently reduced. Other requirements and capacities can further limit what is possible.
Set scoring
For one measurement, the raw deviation is the shortfall below the lower bound or the excess above the upper bound.
- Linear:
weight × deviation - Squared:
weight × deviation²
With weight 10, one deviation of 2 costs 20 with Linear and 40 with Squared. Two separate deviations of 1 cost 20 with either function. Squared therefore places more pressure on a large individual miss.
A weight of 0 is allowed. The preference then adds no optimization cost, although Results can still show raw misses.
A distinct minimum applies separately to every selected person. For example, minimum 3 gives a penalty to each person visiting fewer than 3 different groups. It does not automatically maximize the minimum across all people. To also reward total variety, add another distinct preference with a higher minimum and a smaller linear weight. These remain weighted preferences, not hard guarantees or a general lexicographic objective.
Check feasibility
For separate counting, if every selected person must visit each of G groups exactly V times, each person needs at least G × V attended, selected sessions.
Also compare required visits with available places. For 24 students, 4 stations, 4 sessions, and capacity 6 per station:
- required visits:
24 × 4 = 96; - available assignments:
4 × 4 × 6 = 96.
The totals fit. Availability, fixed placements, together/apart requirements, and minimum occupancies can still prevent a zero-deviation result.
Group Visit goals work with Auto, Local improvement, and Annealing improvement. They are not available with Pure SGP solvers.
Read the result
In Results:
- Open Quality and expand Requirements and preferences.
- Expand Group Visits.
- Select Show details on an individual card.
The card header shows scope and weight. Detail rows are shown for misses and include person, group or group set, actual visits, selected sessions, target, and raw visit deviation. Distinct mode reports different groups rather than repeat visits. The Rotation view shows individual distinct counts, their distribution, and scoped preference targets. The detail is raw evidence; the weighted aggregate is part of Cost Score.
Continue after completed sessions
A partial solver Scope with a saved starting result keeps the unselected sessions immutable (see Run and improve a solve). Those assignments remain part of the full scenario and count toward its Group Visits measurements.
If you instead create a new scenario containing only the remaining sessions, previous visits are not present. Set targets to each person’s remaining required destinations.
The Create station-rotation example action on this page creates a separate saved scenario and opens its Group Visits setup. For the full event workflow, see Run station rotations for groups.
Open in GroupMixer
These links use registered product destinations. Result-specific links remain unavailable until a saved result is selected.