use of org.cpsolver.coursett.constraint.InstructorConstraint in project cpsolver by UniTime.
the class InstructorConflict method getBounds.
@Override
public double[] getBounds(Assignment<Lecture, Placement> assignment, Collection<Lecture> variables) {
double[] bounds = new double[] { 0.0, 0.0 };
Set<InstructorConstraint> constraints = new HashSet<InstructorConstraint>();
for (Lecture lect : variables) {
for (InstructorConstraint ic : lect.getInstructorConstraints()) {
if (!constraints.add(ic))
continue;
if (ic instanceof SoftInstructorConstraint)
bounds[1] += ((SoftInstructorConstraint) ic).getWorstConflicts();
}
}
return bounds;
}
use of org.cpsolver.coursett.constraint.InstructorConstraint in project cpsolver by UniTime.
the class InstructorLunchBreak method getInfo.
@Override
public void getInfo(Assignment<Lecture, Placement> assignment, Map<String, String> info) {
Set<String> violatedLunchBreaks = new TreeSet<String>();
int lunchViolations = 0;
for (InstructorConstraint c : ((TimetableModel) getModel()).getInstructorConstraints()) {
String days = "";
CompactInfo compactInfo = ((InstructorLunchBreakContext) getContext(assignment)).getCompactInfo(c);
for (int i = 0; i < Constants.NR_DAYS; i++) {
if (compactInfo.getLunchDayViolations()[i] > 0) {
if (iFullInfo)
days += (days.isEmpty() ? "" : ", ") + compactInfo.getLunchDayViolations()[i] + " × " + Constants.DAY_NAMES_SHORT[i];
lunchViolations += compactInfo.getLunchDayViolations()[i];
}
}
if (iFullInfo && !days.isEmpty())
violatedLunchBreaks.add(c.getName() + ": " + days);
}
if (lunchViolations > 0) {
info.put("Lunch breaks", getPerc(lunchViolations, 0, ((TimetableModel) getModel()).getInstructorConstraints().size() * Constants.NR_DAYS * getWeeks().size()) + "% (" + lunchViolations + ")");
if (iFullInfo && !violatedLunchBreaks.isEmpty()) {
String message = "";
for (String s : violatedLunchBreaks) message += (message.isEmpty() ? "" : "<br>") + s;
info.put("Lunch break violations", message);
}
}
}
use of org.cpsolver.coursett.constraint.InstructorConstraint in project cpsolver by UniTime.
the class InstructorLunchBreak method getValue.
@Override
public double getValue(Assignment<Lecture, Placement> assignment, Placement value, Set<Placement> conflicts) {
double ret = 0.0;
if (value.getTimeLocation().getStartSlot() <= iLunchEnd && value.getTimeLocation().getStartSlot() + value.getTimeLocation().getLength() > iLunchStart) {
InstructorLunchBreakContext context = (InstructorLunchBreakContext) getContext(assignment);
for (InstructorConstraint constraint : value.variable().getInstructorConstraints()) {
InstructorConstraintContext icx = constraint.getContext(assignment);
CompactInfo compactInfo = context.getCompactInfo(constraint);
for (int i = 0; i < Constants.NR_DAYS; i++) {
// checks only days affected by the placement
if ((value.getTimeLocation().getDayCode() & Constants.DAY_CODES[i]) != 0) {
int currentLunchStartSlot = Constants.SLOTS_PER_DAY * i + iLunchStart;
int currentLunchEndSlot = Constants.SLOTS_PER_DAY * i + iLunchEnd;
int semesterViolations = 0;
for (BitSet week : getWeeks()) {
int maxBreak = 0;
int currentBreak = 0;
for (int slot = currentLunchStartSlot; slot < currentLunchEndSlot; slot++) {
if (isEmpty(icx, slot, week, value)) {
currentBreak++;
if (maxBreak < currentBreak) {
maxBreak = currentBreak;
}
} else {
currentBreak = 0;
}
}
if (maxBreak < iLunchLength) {
semesterViolations++;
}
}
// add the difference to the result
ret += semesterViolations - compactInfo.getLunchDayViolations()[i];
}
}
}
}
return ret;
}
use of org.cpsolver.coursett.constraint.InstructorConstraint in project cpsolver by UniTime.
the class Test method printSomeStuff.
/**
* Create info.txt with some more information about the problem
* @param solution current solution
* @throws IOException an exception that may be thrown
*/
public static void printSomeStuff(Solution<Lecture, Placement> solution) throws IOException {
TimetableModel model = (TimetableModel) solution.getModel();
Assignment<Lecture, Placement> assignment = solution.getAssignment();
File outDir = new File(model.getProperties().getProperty("General.Output", "."));
PrintWriter pw = new PrintWriter(new FileWriter(outDir.toString() + File.separator + "info.txt"));
PrintWriter pwi = new PrintWriter(new FileWriter(outDir.toString() + File.separator + "info.csv"));
String name = new File(model.getProperties().getProperty("General.Input")).getName();
pwi.println("Instance," + name.substring(0, name.lastIndexOf('.')));
pw.println("Solution info: " + ToolBox.dict2string(solution.getInfo(), 1));
pw.println("Bounds: " + ToolBox.dict2string(model.getBounds(assignment), 1));
Map<String, String> info = solution.getInfo();
for (String key : new TreeSet<String>(info.keySet())) {
if (key.equals("Memory usage"))
continue;
if (key.equals("Iteration"))
continue;
if (key.equals("Time"))
continue;
String value = info.get(key);
if (value.indexOf(' ') > 0)
value = value.substring(0, value.indexOf(' '));
pwi.println(key + "," + value);
}
printRoomInfo(pw, model, assignment);
printClassInfo(pw, model);
long nrValues = 0;
long nrTimes = 0;
long nrRooms = 0;
double totalMaxNormTimePref = 0.0;
double totalMinNormTimePref = 0.0;
double totalNormTimePref = 0.0;
int totalMaxRoomPref = 0;
int totalMinRoomPref = 0;
int totalRoomPref = 0;
long nrStudentEnrls = 0;
long nrInevitableStudentConflicts = 0;
long nrJenrls = 0;
int nrHalfHours = 0;
int nrMeetings = 0;
int totalMinLimit = 0;
int totalMaxLimit = 0;
long nrReqRooms = 0;
int nrSingleValueVariables = 0;
int nrSingleTimeVariables = 0;
int nrSingleRoomVariables = 0;
long totalAvailableMinRoomSize = 0;
long totalAvailableMaxRoomSize = 0;
long totalRoomSize = 0;
long nrOneOrMoreRoomVariables = 0;
long nrOneRoomVariables = 0;
HashSet<Student> students = new HashSet<Student>();
HashSet<Long> offerings = new HashSet<Long>();
HashSet<Long> configs = new HashSet<Long>();
HashSet<Long> subparts = new HashSet<Long>();
int[] sizeLimits = new int[] { 0, 25, 50, 75, 100, 150, 200, 400 };
int[] nrRoomsOfSize = new int[sizeLimits.length];
int[] minRoomOfSize = new int[sizeLimits.length];
int[] maxRoomOfSize = new int[sizeLimits.length];
int[] totalUsedSlots = new int[sizeLimits.length];
int[] totalUsedSeats = new int[sizeLimits.length];
int[] totalUsedSeats2 = new int[sizeLimits.length];
int firstDaySlot = model.getProperties().getPropertyInt("General.FirstDaySlot", Constants.DAY_SLOTS_FIRST);
int lastDaySlot = model.getProperties().getPropertyInt("General.LastDaySlot", Constants.DAY_SLOTS_LAST);
int firstWorkDay = model.getProperties().getPropertyInt("General.FirstWorkDay", 0);
int lastWorkDay = model.getProperties().getPropertyInt("General.LastWorkDay", Constants.NR_DAYS_WEEK - 1);
if (lastWorkDay < firstWorkDay)
lastWorkDay += 7;
for (Lecture lect : model.variables()) {
if (lect.getConfiguration() != null) {
offerings.add(lect.getConfiguration().getOfferingId());
configs.add(lect.getConfiguration().getConfigId());
}
subparts.add(lect.getSchedulingSubpartId());
nrStudentEnrls += (lect.students() == null ? 0 : lect.students().size());
students.addAll(lect.students());
nrValues += lect.values(solution.getAssignment()).size();
nrReqRooms += lect.getNrRooms();
for (RoomLocation room : lect.roomLocations()) if (room.getPreference() < Constants.sPreferenceLevelProhibited / 2)
nrRooms++;
for (TimeLocation time : lect.timeLocations()) if (time.getPreference() < Constants.sPreferenceLevelProhibited / 2)
nrTimes++;
totalMinLimit += lect.minClassLimit();
totalMaxLimit += lect.maxClassLimit();
if (!lect.values(solution.getAssignment()).isEmpty()) {
Placement p = lect.values(solution.getAssignment()).get(0);
nrMeetings += p.getTimeLocation().getNrMeetings();
nrHalfHours += p.getTimeLocation().getNrMeetings() * p.getTimeLocation().getNrSlotsPerMeeting();
totalMaxNormTimePref += lect.getMinMaxTimePreference()[1];
totalMinNormTimePref += lect.getMinMaxTimePreference()[0];
totalNormTimePref += Math.abs(lect.getMinMaxTimePreference()[1] - lect.getMinMaxTimePreference()[0]);
totalMaxRoomPref += lect.getMinMaxRoomPreference()[1];
totalMinRoomPref += lect.getMinMaxRoomPreference()[0];
totalRoomPref += Math.abs(lect.getMinMaxRoomPreference()[1] - lect.getMinMaxRoomPreference()[0]);
TimeLocation time = p.getTimeLocation();
boolean hasRoomConstraint = false;
for (RoomLocation roomLocation : lect.roomLocations()) {
if (roomLocation.getRoomConstraint().getConstraint())
hasRoomConstraint = true;
}
if (hasRoomConstraint && lect.getNrRooms() > 0) {
for (int d = firstWorkDay; d <= lastWorkDay; d++) {
if ((time.getDayCode() & Constants.DAY_CODES[d % 7]) == 0)
continue;
for (int t = Math.max(time.getStartSlot(), firstDaySlot); t <= Math.min(time.getStartSlot() + time.getLength() - 1, lastDaySlot); t++) {
for (int l = 0; l < sizeLimits.length; l++) {
if (sizeLimits[l] <= lect.minRoomSize()) {
totalUsedSlots[l] += lect.getNrRooms();
totalUsedSeats[l] += lect.classLimit(assignment);
totalUsedSeats2[l] += lect.minRoomSize() * lect.getNrRooms();
}
}
}
}
}
}
if (lect.values(solution.getAssignment()).size() == 1) {
nrSingleValueVariables++;
}
if (lect.timeLocations().size() == 1) {
nrSingleTimeVariables++;
}
if (lect.roomLocations().size() == 1) {
nrSingleRoomVariables++;
}
if (lect.getNrRooms() == 1) {
nrOneRoomVariables++;
}
if (lect.getNrRooms() > 0) {
nrOneOrMoreRoomVariables++;
}
if (!lect.roomLocations().isEmpty()) {
int minRoomSize = Integer.MAX_VALUE;
int maxRoomSize = Integer.MIN_VALUE;
for (RoomLocation rl : lect.roomLocations()) {
minRoomSize = Math.min(minRoomSize, rl.getRoomSize());
maxRoomSize = Math.max(maxRoomSize, rl.getRoomSize());
totalRoomSize += rl.getRoomSize();
}
totalAvailableMinRoomSize += minRoomSize;
totalAvailableMaxRoomSize += maxRoomSize;
}
}
for (JenrlConstraint jenrl : model.getJenrlConstraints()) {
nrJenrls += jenrl.getJenrl();
if ((jenrl.first()).timeLocations().size() == 1 && (jenrl.second()).timeLocations().size() == 1) {
TimeLocation t1 = jenrl.first().timeLocations().get(0);
TimeLocation t2 = jenrl.second().timeLocations().get(0);
if (t1.hasIntersection(t2)) {
nrInevitableStudentConflicts += jenrl.getJenrl();
pw.println("Inevitable " + jenrl.getJenrl() + " student conflicts between " + jenrl.first() + " " + t1 + " and " + jenrl.second() + " " + t2);
} else if (jenrl.first().values(solution.getAssignment()).size() == 1 && jenrl.second().values(solution.getAssignment()).size() == 1) {
Placement p1 = jenrl.first().values(solution.getAssignment()).get(0);
Placement p2 = jenrl.second().values(solution.getAssignment()).get(0);
if (JenrlConstraint.isInConflict(p1, p2, ((TimetableModel) p1.variable().getModel()).getDistanceMetric(), ((TimetableModel) p1.variable().getModel()).getStudentWorkDayLimit())) {
nrInevitableStudentConflicts += jenrl.getJenrl();
pw.println("Inevitable " + jenrl.getJenrl() + (p1.getTimeLocation().hasIntersection(p2.getTimeLocation()) ? "" : " distance") + " student conflicts between " + p1 + " and " + p2);
}
}
}
}
int totalCommitedPlacements = 0;
for (Student student : students) {
if (student.getCommitedPlacements() != null)
totalCommitedPlacements += student.getCommitedPlacements().size();
}
pw.println("Total number of classes: " + model.variables().size());
pwi.println("Number of classes," + model.variables().size());
pw.println("Total number of instructional offerings: " + offerings.size() + " (" + sDoubleFormat.format(100.0 * offerings.size() / model.variables().size()) + "%)");
// pwi.println("Number of instructional offerings,"+offerings.size());
pw.println("Total number of configurations: " + configs.size() + " (" + sDoubleFormat.format(100.0 * configs.size() / model.variables().size()) + "%)");
pw.println("Total number of scheduling subparts: " + subparts.size() + " (" + sDoubleFormat.format(100.0 * subparts.size() / model.variables().size()) + "%)");
// pwi.println("Number of scheduling subparts,"+subparts.size());
pw.println("Average number classes per subpart: " + sDoubleFormat.format(1.0 * model.variables().size() / subparts.size()));
pwi.println("Avg. classes per instruction," + sDoubleFormat.format(1.0 * model.variables().size() / subparts.size()));
pw.println("Average number classes per config: " + sDoubleFormat.format(1.0 * model.variables().size() / configs.size()));
pw.println("Average number classes per offering: " + sDoubleFormat.format(1.0 * model.variables().size() / offerings.size()));
pw.println("Total number of classes with only one value: " + nrSingleValueVariables + " (" + sDoubleFormat.format(100.0 * nrSingleValueVariables / model.variables().size()) + "%)");
pw.println("Total number of classes with only one time: " + nrSingleTimeVariables + " (" + sDoubleFormat.format(100.0 * nrSingleTimeVariables / model.variables().size()) + "%)");
pw.println("Total number of classes with only one room: " + nrSingleRoomVariables + " (" + sDoubleFormat.format(100.0 * nrSingleRoomVariables / model.variables().size()) + "%)");
pwi.println("Classes with single value," + nrSingleValueVariables);
// pwi.println("Classes with only one time/room,"+nrSingleTimeVariables+"/"+nrSingleRoomVariables);
pw.println("Total number of classes requesting no room: " + (model.variables().size() - nrOneOrMoreRoomVariables) + " (" + sDoubleFormat.format(100.0 * (model.variables().size() - nrOneOrMoreRoomVariables) / model.variables().size()) + "%)");
pw.println("Total number of classes requesting one room: " + nrOneRoomVariables + " (" + sDoubleFormat.format(100.0 * nrOneRoomVariables / model.variables().size()) + "%)");
pw.println("Total number of classes requesting one or more rooms: " + nrOneOrMoreRoomVariables + " (" + sDoubleFormat.format(100.0 * nrOneOrMoreRoomVariables / model.variables().size()) + "%)");
// pwi.println("% classes requesting no room,"+sDoubleFormat.format(100.0*(model.variables().size()-nrOneOrMoreRoomVariables)/model.variables().size())+"%");
// pwi.println("% classes requesting one room,"+sDoubleFormat.format(100.0*nrOneRoomVariables/model.variables().size())+"%");
// pwi.println("% classes requesting two or more rooms,"+sDoubleFormat.format(100.0*(nrOneOrMoreRoomVariables-nrOneRoomVariables)/model.variables().size())+"%");
pw.println("Average number of requested rooms: " + sDoubleFormat.format(1.0 * nrReqRooms / model.variables().size()));
pw.println("Average minimal class limit: " + sDoubleFormat.format(1.0 * totalMinLimit / model.variables().size()));
pw.println("Average maximal class limit: " + sDoubleFormat.format(1.0 * totalMaxLimit / model.variables().size()));
// pwi.println("Average class limit,"+sDoubleFormat.format(1.0*(totalMinLimit+totalMaxLimit)/(2*model.variables().size())));
pw.println("Average number of placements: " + sDoubleFormat.format(1.0 * nrValues / model.variables().size()));
// pwi.println("Average domain size,"+sDoubleFormat.format(1.0*nrValues/model.variables().size()));
pwi.println("Avg. domain size," + sDoubleFormat.format(1.0 * nrValues / model.variables().size()));
pw.println("Average number of time locations: " + sDoubleFormat.format(1.0 * nrTimes / model.variables().size()));
pwi.println("Avg. number of avail. times/rooms," + sDoubleFormat.format(1.0 * nrTimes / model.variables().size()) + "/" + sDoubleFormat.format(1.0 * nrRooms / model.variables().size()));
pw.println("Average number of room locations: " + sDoubleFormat.format(1.0 * nrRooms / model.variables().size()));
pw.println("Average minimal requested room size: " + sDoubleFormat.format(1.0 * totalAvailableMinRoomSize / nrOneOrMoreRoomVariables));
pw.println("Average maximal requested room size: " + sDoubleFormat.format(1.0 * totalAvailableMaxRoomSize / nrOneOrMoreRoomVariables));
pw.println("Average requested room sizes: " + sDoubleFormat.format(1.0 * totalRoomSize / nrRooms));
pwi.println("Average requested room size," + sDoubleFormat.format(1.0 * totalRoomSize / nrRooms));
pw.println("Average maximum normalized time preference: " + sDoubleFormat.format(totalMaxNormTimePref / model.variables().size()));
pw.println("Average minimum normalized time preference: " + sDoubleFormat.format(totalMinNormTimePref / model.variables().size()));
pw.println("Average normalized time preference," + sDoubleFormat.format(totalNormTimePref / model.variables().size()));
pw.println("Average maximum room preferences: " + sDoubleFormat.format(1.0 * totalMaxRoomPref / nrOneOrMoreRoomVariables));
pw.println("Average minimum room preferences: " + sDoubleFormat.format(1.0 * totalMinRoomPref / nrOneOrMoreRoomVariables));
pw.println("Average room preferences," + sDoubleFormat.format(1.0 * totalRoomPref / nrOneOrMoreRoomVariables));
pw.println("Total number of students:" + students.size());
pwi.println("Number of students," + students.size());
pwi.println("Number of inevitable student conflicts," + nrInevitableStudentConflicts);
pw.println("Total amount of student enrollments: " + nrStudentEnrls);
pwi.println("Number of student enrollments," + nrStudentEnrls);
pw.println("Total amount of joined enrollments: " + nrJenrls);
pwi.println("Number of joint student enrollments," + nrJenrls);
pw.println("Average number of students: " + sDoubleFormat.format(1.0 * students.size() / model.variables().size()));
pw.println("Average number of enrollemnts (per student): " + sDoubleFormat.format(1.0 * nrStudentEnrls / students.size()));
pwi.println("Avg. number of classes per student," + sDoubleFormat.format(1.0 * nrStudentEnrls / students.size()));
pwi.println("Avg. number of committed classes per student," + sDoubleFormat.format(1.0 * totalCommitedPlacements / students.size()));
pw.println("Total amount of inevitable student conflicts: " + nrInevitableStudentConflicts + " (" + sDoubleFormat.format(100.0 * nrInevitableStudentConflicts / nrStudentEnrls) + "%)");
pw.println("Average number of meetings (per class): " + sDoubleFormat.format(1.0 * nrMeetings / model.variables().size()));
pw.println("Average number of hours per class: " + sDoubleFormat.format(1.0 * nrHalfHours / model.variables().size() / 12.0));
pwi.println("Avg. number of meetings per class," + sDoubleFormat.format(1.0 * nrMeetings / model.variables().size()));
pwi.println("Avg. number of hours per class," + sDoubleFormat.format(1.0 * nrHalfHours / model.variables().size() / 12.0));
int minRoomSize = Integer.MAX_VALUE;
int maxRoomSize = Integer.MIN_VALUE;
int nrDistancePairs = 0;
double maxRoomDistance = Double.MIN_VALUE;
double totalRoomDistance = 0.0;
int[] totalAvailableSlots = new int[sizeLimits.length];
int[] totalAvailableSeats = new int[sizeLimits.length];
int nrOfRooms = 0;
totalRoomSize = 0;
for (RoomConstraint rc : model.getRoomConstraints()) {
if (rc.variables().isEmpty())
continue;
nrOfRooms++;
minRoomSize = Math.min(minRoomSize, rc.getCapacity());
maxRoomSize = Math.max(maxRoomSize, rc.getCapacity());
for (int l = 0; l < sizeLimits.length; l++) {
if (sizeLimits[l] <= rc.getCapacity() && (l + 1 == sizeLimits.length || rc.getCapacity() < sizeLimits[l + 1])) {
nrRoomsOfSize[l]++;
if (minRoomOfSize[l] == 0)
minRoomOfSize[l] = rc.getCapacity();
else
minRoomOfSize[l] = Math.min(minRoomOfSize[l], rc.getCapacity());
if (maxRoomOfSize[l] == 0)
maxRoomOfSize[l] = rc.getCapacity();
else
maxRoomOfSize[l] = Math.max(maxRoomOfSize[l], rc.getCapacity());
}
}
totalRoomSize += rc.getCapacity();
if (rc.getPosX() != null && rc.getPosY() != null) {
for (RoomConstraint rc2 : model.getRoomConstraints()) {
if (rc2.getResourceId().compareTo(rc.getResourceId()) > 0 && rc2.getPosX() != null && rc2.getPosY() != null) {
double distance = ((TimetableModel) solution.getModel()).getDistanceMetric().getDistanceInMinutes(rc.getId(), rc.getPosX(), rc.getPosY(), rc2.getId(), rc2.getPosX(), rc2.getPosY());
totalRoomDistance += distance;
nrDistancePairs++;
maxRoomDistance = Math.max(maxRoomDistance, distance);
}
}
}
for (int d = firstWorkDay; d <= lastWorkDay; d++) {
for (int t = firstDaySlot; t <= lastDaySlot; t++) {
if (rc.isAvailable((d % 7) * Constants.SLOTS_PER_DAY + t)) {
for (int l = 0; l < sizeLimits.length; l++) {
if (sizeLimits[l] <= rc.getCapacity()) {
totalAvailableSlots[l]++;
totalAvailableSeats[l] += rc.getCapacity();
}
}
}
}
}
}
pw.println("Total number of rooms: " + nrOfRooms);
pwi.println("Number of rooms," + nrOfRooms);
pw.println("Minimal room size: " + minRoomSize);
pw.println("Maximal room size: " + maxRoomSize);
pwi.println("Room size min/max," + minRoomSize + "/" + maxRoomSize);
pw.println("Average room size: " + sDoubleFormat.format(1.0 * totalRoomSize / model.getRoomConstraints().size()));
pw.println("Maximal distance between two rooms: " + sDoubleFormat.format(maxRoomDistance));
pw.println("Average distance between two rooms: " + sDoubleFormat.format(totalRoomDistance / nrDistancePairs));
pwi.println("Average distance between two rooms [min]," + sDoubleFormat.format(totalRoomDistance / nrDistancePairs));
pwi.println("Maximal distance between two rooms [min]," + sDoubleFormat.format(maxRoomDistance));
for (int l = 0; l < sizeLimits.length; l++) {
// sizeLimits.length;l++) {
pwi.println("\"Room frequency (size>=" + sizeLimits[l] + ", used/avaiable times)\"," + sDoubleFormat.format(100.0 * totalUsedSlots[l] / totalAvailableSlots[l]) + "%");
pwi.println("\"Room utilization (size>=" + sizeLimits[l] + ", used/available seats)\"," + sDoubleFormat.format(100.0 * totalUsedSeats[l] / totalAvailableSeats[l]) + "%");
pwi.println("\"Number of rooms (size>=" + sizeLimits[l] + ")\"," + nrRoomsOfSize[l]);
pwi.println("\"Min/max room size (size>=" + sizeLimits[l] + ")\"," + minRoomOfSize[l] + "-" + maxRoomOfSize[l]);
// pwi.println("\"Room utilization (size>="+sizeLimits[l]+", minRoomSize)\","+sDoubleFormat.format(100.0*totalUsedSeats2[l]/totalAvailableSeats[l])+"%");
}
pw.println("Average hours available: " + sDoubleFormat.format(1.0 * totalAvailableSlots[0] / nrOfRooms / 12.0));
int totalInstructedClasses = 0;
for (InstructorConstraint ic : model.getInstructorConstraints()) {
totalInstructedClasses += ic.variables().size();
}
pw.println("Total number of instructors: " + model.getInstructorConstraints().size());
pwi.println("Number of instructors," + model.getInstructorConstraints().size());
pw.println("Total class-instructor assignments: " + totalInstructedClasses + " (" + sDoubleFormat.format(100.0 * totalInstructedClasses / model.variables().size()) + "%)");
pwi.println("Number of class-instructor assignments," + totalInstructedClasses);
pw.println("Average classes per instructor: " + sDoubleFormat.format(1.0 * totalInstructedClasses / model.getInstructorConstraints().size()));
pwi.println("Average classes per instructor," + sDoubleFormat.format(1.0 * totalInstructedClasses / model.getInstructorConstraints().size()));
// pw.println("Average hours available: "+sDoubleFormat.format(1.0*totalAvailableSlots/model.getInstructorConstraints().size()/12.0));
// pwi.println("Instructor availability [h],"+sDoubleFormat.format(1.0*totalAvailableSlots/model.getInstructorConstraints().size()/12.0));
int nrGroupConstraints = model.getGroupConstraints().size() + model.getSpreadConstraints().size();
int nrHardGroupConstraints = 0;
int nrVarsInGroupConstraints = 0;
for (GroupConstraint gc : model.getGroupConstraints()) {
if (gc.isHard())
nrHardGroupConstraints++;
nrVarsInGroupConstraints += gc.variables().size();
}
for (SpreadConstraint sc : model.getSpreadConstraints()) {
nrVarsInGroupConstraints += sc.variables().size();
}
pw.println("Total number of group constraints: " + nrGroupConstraints + " (" + sDoubleFormat.format(100.0 * nrGroupConstraints / model.variables().size()) + "%)");
// pwi.println("Number of group constraints,"+nrGroupConstraints);
pw.println("Total number of hard group constraints: " + nrHardGroupConstraints + " (" + sDoubleFormat.format(100.0 * nrHardGroupConstraints / model.variables().size()) + "%)");
// pwi.println("Number of hard group constraints,"+nrHardGroupConstraints);
pw.println("Average classes per group constraint: " + sDoubleFormat.format(1.0 * nrVarsInGroupConstraints / nrGroupConstraints));
// pwi.println("Average classes per group constraint,"+sDoubleFormat.format(1.0*nrVarsInGroupConstraints/nrGroupConstraints));
pwi.println("Avg. number distribution constraints per class," + sDoubleFormat.format(1.0 * nrVarsInGroupConstraints / model.variables().size()));
pwi.println("Joint enrollment constraints," + model.getJenrlConstraints().size());
pw.flush();
pw.close();
pwi.flush();
pwi.close();
}
use of org.cpsolver.coursett.constraint.InstructorConstraint in project cpsolver by UniTime.
the class Test method saveOutputCSV.
public static void saveOutputCSV(Solution<Lecture, Placement> s, File file) {
try {
DecimalFormat dx = new DecimalFormat("000");
PrintWriter w = new PrintWriter(new FileWriter(file));
TimetableModel m = (TimetableModel) s.getModel();
int firstDaySlot = m.getProperties().getPropertyInt("General.FirstDaySlot", Constants.DAY_SLOTS_FIRST);
int lastDaySlot = m.getProperties().getPropertyInt("General.LastDaySlot", Constants.DAY_SLOTS_LAST);
int firstWorkDay = m.getProperties().getPropertyInt("General.FirstWorkDay", 0);
int lastWorkDay = m.getProperties().getPropertyInt("General.LastWorkDay", Constants.NR_DAYS_WEEK - 1);
if (lastWorkDay < firstWorkDay)
lastWorkDay += 7;
Assignment<Lecture, Placement> a = s.getAssignment();
int idx = 1;
w.println("000." + dx.format(idx++) + " Assigned variables," + a.nrAssignedVariables());
w.println("000." + dx.format(idx++) + " Time [sec]," + sDoubleFormat.format(s.getBestTime()));
w.println("000." + dx.format(idx++) + " Hard student conflicts," + Math.round(m.getCriterion(StudentHardConflict.class).getValue(a)));
if (m.getProperties().getPropertyBoolean("General.UseDistanceConstraints", true))
w.println("000." + dx.format(idx++) + " Distance student conf.," + Math.round(m.getCriterion(StudentDistanceConflict.class).getValue(a)));
w.println("000." + dx.format(idx++) + " Student conflicts," + Math.round(m.getCriterion(StudentConflict.class).getValue(a)));
w.println("000." + dx.format(idx++) + " Committed student conflicts," + Math.round(m.getCriterion(StudentCommittedConflict.class).getValue(a)));
w.println("000." + dx.format(idx++) + " All Student conflicts," + Math.round(m.getCriterion(StudentConflict.class).getValue(a) + m.getCriterion(StudentCommittedConflict.class).getValue(a)));
w.println("000." + dx.format(idx++) + " Time preferences," + sDoubleFormat.format(m.getCriterion(TimePreferences.class).getValue(a)));
w.println("000." + dx.format(idx++) + " Room preferences," + Math.round(m.getCriterion(RoomPreferences.class).getValue(a)));
w.println("000." + dx.format(idx++) + " Useless half-hours," + Math.round(m.getCriterion(UselessHalfHours.class).getValue(a)));
w.println("000." + dx.format(idx++) + " Broken time patterns," + Math.round(m.getCriterion(BrokenTimePatterns.class).getValue(a)));
w.println("000." + dx.format(idx++) + " Too big room," + Math.round(m.getCriterion(TooBigRooms.class).getValue(a)));
w.println("000." + dx.format(idx++) + " Distribution preferences," + sDoubleFormat.format(m.getCriterion(DistributionPreferences.class).getValue(a)));
if (m.getProperties().getPropertyBoolean("General.UseDistanceConstraints", true))
w.println("000." + dx.format(idx++) + " Back-to-back instructor pref.," + Math.round(m.getCriterion(BackToBackInstructorPreferences.class).getValue(a)));
if (m.getProperties().getPropertyBoolean("General.DeptBalancing", true)) {
w.println("000." + dx.format(idx++) + " Dept. balancing penalty," + sDoubleFormat.format(m.getCriterion(DepartmentBalancingPenalty.class).getValue(a)));
}
w.println("000." + dx.format(idx++) + " Same subpart balancing penalty," + sDoubleFormat.format(m.getCriterion(SameSubpartBalancingPenalty.class).getValue(a)));
if (m.getProperties().getPropertyBoolean("General.MPP", false)) {
Map<String, Double> mppInfo = ((UniversalPerturbationsCounter) ((Perturbations) m.getCriterion(Perturbations.class)).getPerturbationsCounter()).getCompactInfo(a, m, false, false);
int pidx = 51;
w.println("000." + dx.format(pidx++) + " Perturbation penalty," + sDoubleFormat.format(m.getCriterion(Perturbations.class).getValue(a)));
w.println("000." + dx.format(pidx++) + " Additional perturbations," + m.perturbVariables(a).size());
int nrPert = 0, nrStudentPert = 0;
for (Lecture lecture : m.variables()) {
if (lecture.getInitialAssignment() != null)
continue;
nrPert++;
nrStudentPert += lecture.classLimit(a);
}
w.println("000." + dx.format(pidx++) + " Given perturbations," + nrPert);
w.println("000." + dx.format(pidx++) + " Given student perturbations," + nrStudentPert);
for (String key : new TreeSet<String>(mppInfo.keySet())) {
Double value = mppInfo.get(key);
w.println("000." + dx.format(pidx++) + " " + key + "," + sDoubleFormat.format(value));
}
}
HashSet<Student> students = new HashSet<Student>();
int enrls = 0;
int minRoomPref = 0, maxRoomPref = 0;
int minGrPref = 0, maxGrPref = 0;
int minTimePref = 0, maxTimePref = 0;
int worstInstrPref = 0;
HashSet<Constraint<Lecture, Placement>> used = new HashSet<Constraint<Lecture, Placement>>();
for (Lecture lecture : m.variables()) {
enrls += (lecture.students() == null ? 0 : lecture.students().size());
students.addAll(lecture.students());
int[] minMaxRoomPref = lecture.getMinMaxRoomPreference();
maxRoomPref += minMaxRoomPref[1] - minMaxRoomPref[0];
double[] minMaxTimePref = lecture.getMinMaxTimePreference();
maxTimePref += minMaxTimePref[1] - minMaxTimePref[0];
for (Constraint<Lecture, Placement> c : lecture.constraints()) {
if (!used.add(c))
continue;
if (c instanceof InstructorConstraint) {
InstructorConstraint ic = (InstructorConstraint) c;
worstInstrPref += ic.getWorstPreference();
}
if (c instanceof GroupConstraint) {
GroupConstraint gc = (GroupConstraint) c;
if (gc.isHard())
continue;
maxGrPref += Math.abs(gc.getPreference()) * (1 + (gc.variables().size() * (gc.variables().size() - 1)) / 2);
}
}
}
int totalCommitedPlacements = 0;
for (Student student : students) {
if (student.getCommitedPlacements() != null)
totalCommitedPlacements += student.getCommitedPlacements().size();
}
HashMap<Long, List<Lecture>> subs = new HashMap<Long, List<Lecture>>();
for (Lecture lecture : m.variables()) {
if (lecture.isCommitted() || lecture.getScheduler() == null)
continue;
List<Lecture> vars = subs.get(lecture.getScheduler());
if (vars == null) {
vars = new ArrayList<Lecture>();
subs.put(lecture.getScheduler(), vars);
}
vars.add(lecture);
}
int bidx = 101;
w.println("000." + dx.format(bidx++) + " Assigned variables max," + m.variables().size());
w.println("000." + dx.format(bidx++) + " Student enrollments," + enrls);
w.println("000." + dx.format(bidx++) + " Student commited enrollments," + totalCommitedPlacements);
w.println("000." + dx.format(bidx++) + " All student enrollments," + (enrls + totalCommitedPlacements));
w.println("000." + dx.format(bidx++) + " Time preferences min," + minTimePref);
w.println("000." + dx.format(bidx++) + " Time preferences max," + maxTimePref);
w.println("000." + dx.format(bidx++) + " Room preferences min," + minRoomPref);
w.println("000." + dx.format(bidx++) + " Room preferences max," + maxRoomPref);
w.println("000." + dx.format(bidx++) + " Useless half-hours max," + (Constants.sPreferenceLevelStronglyDiscouraged * m.getRoomConstraints().size() * (lastDaySlot - firstDaySlot + 1) * (lastWorkDay - firstWorkDay + 1)));
w.println("000." + dx.format(bidx++) + " Too big room max," + (Constants.sPreferenceLevelStronglyDiscouraged * m.variables().size()));
w.println("000." + dx.format(bidx++) + " Distribution preferences min," + minGrPref);
w.println("000." + dx.format(bidx++) + " Distribution preferences max," + maxGrPref);
w.println("000." + dx.format(bidx++) + " Back-to-back instructor pref max," + worstInstrPref);
TooBigRooms tbr = (TooBigRooms) m.getCriterion(TooBigRooms.class);
for (Long scheduler : new TreeSet<Long>(subs.keySet())) {
List<Lecture> vars = subs.get(scheduler);
idx = 001;
bidx = 101;
int nrAssg = 0;
enrls = 0;
int roomPref = 0;
minRoomPref = 0;
maxRoomPref = 0;
double timePref = 0;
minTimePref = 0;
maxTimePref = 0;
double grPref = 0;
minGrPref = 0;
maxGrPref = 0;
long allSC = 0, hardSC = 0, distSC = 0;
int instPref = 0;
worstInstrPref = 0;
int spreadPen = 0, deptSpreadPen = 0;
int tooBigRooms = 0;
int rcs = 0, uselessSlots = 0;
used = new HashSet<Constraint<Lecture, Placement>>();
for (Lecture lecture : vars) {
if (lecture.isCommitted())
continue;
enrls += lecture.students().size();
Placement placement = a.getValue(lecture);
if (placement != null) {
nrAssg++;
}
int[] minMaxRoomPref = lecture.getMinMaxRoomPreference();
minRoomPref += minMaxRoomPref[0];
maxRoomPref += minMaxRoomPref[1];
double[] minMaxTimePref = lecture.getMinMaxTimePreference();
minTimePref += minMaxTimePref[0];
maxTimePref += minMaxTimePref[1];
if (placement != null) {
roomPref += placement.getRoomPreference();
timePref += placement.getTimeLocation().getNormalizedPreference();
if (tbr != null)
tooBigRooms += tbr.getPreference(placement);
}
for (Constraint<Lecture, Placement> c : lecture.constraints()) {
if (!used.add(c))
continue;
if (c instanceof InstructorConstraint) {
InstructorConstraint ic = (InstructorConstraint) c;
instPref += ic.getPreference(a);
worstInstrPref += ic.getWorstPreference();
}
if (c instanceof DepartmentSpreadConstraint) {
DepartmentSpreadConstraint dsc = (DepartmentSpreadConstraint) c;
deptSpreadPen += dsc.getPenalty(a);
} else if (c instanceof SpreadConstraint) {
SpreadConstraint sc = (SpreadConstraint) c;
spreadPen += sc.getPenalty(a);
}
if (c instanceof GroupConstraint) {
GroupConstraint gc = (GroupConstraint) c;
if (gc.isHard())
continue;
minGrPref -= Math.abs(gc.getPreference());
maxGrPref += 0;
grPref += Math.min(0, gc.getCurrentPreference(a));
// minGrPref += Math.min(gc.getPreference(), 0);
// maxGrPref += Math.max(gc.getPreference(), 0);
// grPref += gc.getCurrentPreference();
}
if (c instanceof JenrlConstraint) {
JenrlConstraint jc = (JenrlConstraint) c;
if (!jc.isInConflict(a) || !jc.isOfTheSameProblem())
continue;
Lecture l1 = jc.first();
Lecture l2 = jc.second();
allSC += jc.getJenrl();
if (l1.areStudentConflictsHard(l2))
hardSC += jc.getJenrl();
Placement p1 = a.getValue(l1);
Placement p2 = a.getValue(l2);
if (!p1.getTimeLocation().hasIntersection(p2.getTimeLocation()))
distSC += jc.getJenrl();
}
if (c instanceof RoomConstraint) {
RoomConstraint rc = (RoomConstraint) c;
uselessSlots += UselessHalfHours.countUselessSlotsHalfHours(rc.getContext(a)) + BrokenTimePatterns.countUselessSlotsBrokenTimePatterns(rc.getContext(a));
rcs++;
}
}
}
w.println(dx.format(scheduler) + "." + dx.format(idx++) + " Assigned variables," + nrAssg);
w.println(dx.format(scheduler) + "." + dx.format(bidx++) + " Assigned variables max," + vars.size());
w.println(dx.format(scheduler) + "." + dx.format(idx++) + " Hard student conflicts," + hardSC);
w.println(dx.format(scheduler) + "." + dx.format(bidx++) + " Student enrollments," + enrls);
if (m.getProperties().getPropertyBoolean("General.UseDistanceConstraints", true))
w.println(dx.format(scheduler) + "." + dx.format(idx++) + " Distance student conf.," + distSC);
w.println(dx.format(scheduler) + "." + dx.format(idx++) + " Student conflicts," + allSC);
w.println(dx.format(scheduler) + "." + dx.format(idx++) + " Time preferences," + timePref);
w.println(dx.format(scheduler) + "." + dx.format(bidx++) + " Time preferences min," + minTimePref);
w.println(dx.format(scheduler) + "." + dx.format(bidx++) + " Time preferences max," + maxTimePref);
w.println(dx.format(scheduler) + "." + dx.format(idx++) + " Room preferences," + roomPref);
w.println(dx.format(scheduler) + "." + dx.format(bidx++) + " Room preferences min," + minRoomPref);
w.println(dx.format(scheduler) + "." + dx.format(bidx++) + " Room preferences max," + maxRoomPref);
w.println(dx.format(scheduler) + "." + dx.format(idx++) + " Useless half-hours," + uselessSlots);
w.println(dx.format(scheduler) + "." + dx.format(bidx++) + " Useless half-hours max," + (Constants.sPreferenceLevelStronglyDiscouraged * rcs * (lastDaySlot - firstDaySlot + 1) * (lastWorkDay - firstWorkDay + 1)));
w.println(dx.format(scheduler) + "." + dx.format(idx++) + " Too big room," + tooBigRooms);
w.println(dx.format(scheduler) + "." + dx.format(bidx++) + " Too big room max," + (Constants.sPreferenceLevelStronglyDiscouraged * vars.size()));
w.println(dx.format(scheduler) + "." + dx.format(idx++) + " Distribution preferences," + grPref);
w.println(dx.format(scheduler) + "." + dx.format(bidx++) + " Distribution preferences min," + minGrPref);
w.println(dx.format(scheduler) + "." + dx.format(bidx++) + " Distribution preferences max," + maxGrPref);
if (m.getProperties().getPropertyBoolean("General.UseDistanceConstraints", true))
w.println(dx.format(scheduler) + "." + dx.format(idx++) + " Back-to-back instructor pref," + instPref);
w.println(dx.format(scheduler) + "." + dx.format(bidx++) + " Back-to-back instructor pref max," + worstInstrPref);
if (m.getProperties().getPropertyBoolean("General.DeptBalancing", true)) {
w.println(dx.format(scheduler) + "." + dx.format(idx++) + " Department balancing penalty," + sDoubleFormat.format((deptSpreadPen) / 12.0));
}
w.println(dx.format(scheduler) + "." + dx.format(idx++) + " Same subpart balancing penalty," + sDoubleFormat.format((spreadPen) / 12.0));
}
w.flush();
w.close();
} catch (java.io.IOException io) {
sLogger.error(io.getMessage(), io);
}
}
Aggregations