mmas-3d-no-incentive.ipynb

{
  "cells": [
    {
      "cell_type": "code",
      "execution_count": 79,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-08-21T11:15:56.112887Z",
          "iopub.status.busy": "2022-08-21T11:15:56.112258Z",
          "iopub.status.idle": "2022-08-21T11:15:56.117560Z",
          "shell.execute_reply": "2022-08-21T11:15:56.116340Z",
          "shell.execute_reply.started": "2022-08-21T11:15:56.112850Z"
        },
        "id": "07UBrcaRS2to",
        "trusted": true
      },
      "outputs": [],
      "source": [
        "import numpy as np \n",
        "import pandas as pd "
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 80,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-08-21T11:15:56.127498Z",
          "iopub.status.busy": "2022-08-21T11:15:56.126835Z",
          "iopub.status.idle": "2022-08-21T11:15:56.151248Z",
          "shell.execute_reply": "2022-08-21T11:15:56.150446Z",
          "shell.execute_reply.started": "2022-08-21T11:15:56.127461Z"
        },
        "id": "R171pKX1S2tp",
        "trusted": true
      },
      "outputs": [],
      "source": [
        "students = pd.read_csv(\"data_set/students.csv\")\n",
        "exams = pd.read_csv(\"data_set/exams.csv\")\n",
        "enrolments = pd.read_csv(\"data_set/enrolements.csv\")\n",
        "rooms = pd.read_csv(\"data_set/indexed_rooms.csv\")\n",
        "slots= pd.read_csv(\"data_set/slots.csv\")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 81,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-08-21T11:15:56.155472Z",
          "iopub.status.busy": "2022-08-21T11:15:56.154614Z",
          "iopub.status.idle": "2022-08-21T11:15:56.167616Z",
          "shell.execute_reply": "2022-08-21T11:15:56.166653Z",
          "shell.execute_reply.started": "2022-08-21T11:15:56.155437Z"
        },
        "id": "Q5KSEyk-S2tq",
        "trusted": true
      },
      "outputs": [],
      "source": [
        "class Exam:\n",
        "    def __init__(self, exam_code, name, duration, department_code):\n",
        "        self.code = exam_code\n",
        "        self.name = name\n",
        "        self.duration = duration\n",
        "        self.duration_min = 0\n",
        "        self.dep_code = department_code\n",
        "        self.conflicts = 0\n",
        "        self.exam_conflicts = set()\n",
        "        self.priority = 0\n",
        "        self.slot = -1\n",
        "        self.student_codes = []\n",
        "        self.students = []\n",
        "        self.rooms = []\n",
        "        self.room_assigned = False\n",
        "        self.access_requirement = False\n",
        "        self.students_with_access_needs = []\n",
        "        self.students_assigned = 0\n",
        "\n",
        "class Student:\n",
        "    def __init__(self, student_code, course_code, accessibility_need):\n",
        "        self.code = student_code\n",
        "        self.course_code = course_code\n",
        "        self.accessibility_need = accessibility_need\n",
        "        self.exams = []\n",
        "        self.exam_codes = []\n",
        "\n",
        "class Enrolment:\n",
        "    def __init__(self, student_code, exam_code):\n",
        "        self.student_code = student_code\n",
        "        self.exam_code = exam_code\n",
        "\n",
        "class Room:\n",
        "    def __init__(self, index, room_code, capacity, accessible):\n",
        "        self.index = index\n",
        "        self.code = room_code\n",
        "        self.capacity = capacity\n",
        "        self.status = False\n",
        "        self.accessible = accessible\n",
        "        self.seats_occupied = 0\n",
        "class Slot:\n",
        "    def __init__(self, index, date, day, type):\n",
        "        self.index = index\n",
        "        self.date = date\n",
        "        self.day = day\n",
        "        self.type = type\n",
        "        self.exams = []\n",
        "        self.student_count = 0"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 82,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-08-21T11:15:56.212490Z",
          "iopub.status.busy": "2022-08-21T11:15:56.211851Z",
          "iopub.status.idle": "2022-08-21T11:15:56.221457Z",
          "shell.execute_reply": "2022-08-21T11:15:56.220559Z",
          "shell.execute_reply.started": "2022-08-21T11:15:56.212454Z"
        },
        "id": "kKrr0tN5S2tr",
        "trusted": true
      },
      "outputs": [],
      "source": [
        "exam_objects = [Exam(**kwargs) for kwargs in exams.to_dict(orient='records')]\n",
        "\n",
        "student_objects = [Student(**kwargs) for kwargs in students.to_dict(orient='records')]\n",
        "\n",
        "enrolment_objects = [Enrolment(**kwargs) for kwargs in enrolments.to_dict(orient='records')]\n",
        "\n",
        "room_objects = [Room(**kwargs) for kwargs in rooms.to_dict(orient='records')]\n",
        "\n",
        "slot_objects = [Slot(**kwargs) for kwargs in slots.to_dict(orient='records')]"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 83,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-08-21T11:15:56.265179Z",
          "iopub.status.busy": "2022-08-21T11:15:56.264411Z",
          "iopub.status.idle": "2022-08-21T11:15:56.270114Z",
          "shell.execute_reply": "2022-08-21T11:15:56.269163Z",
          "shell.execute_reply.started": "2022-08-21T11:15:56.265137Z"
        },
        "id": "ZmCefTMUS2tr",
        "trusted": true
      },
      "outputs": [],
      "source": [
        "for i in enrolment_objects:\n",
        "    for j in student_objects:\n",
        "        if j.code == i.student_code:\n",
        "            j.exam_codes.append(i.exam_code)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 84,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-08-21T11:15:56.320147Z",
          "iopub.status.busy": "2022-08-21T11:15:56.319586Z",
          "iopub.status.idle": "2022-08-21T11:15:56.325910Z",
          "shell.execute_reply": "2022-08-21T11:15:56.324870Z",
          "shell.execute_reply.started": "2022-08-21T11:15:56.320107Z"
        },
        "id": "KF9vZAWkS2ts",
        "trusted": true
      },
      "outputs": [],
      "source": [
        "for i in student_objects:\n",
        "    for j in exam_objects:\n",
        "        for k in i.exam_codes:\n",
        "            if k == j.code:\n",
        "                i.exams.append(j)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 85,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-08-21T11:15:56.329582Z",
          "iopub.status.busy": "2022-08-21T11:15:56.328870Z",
          "iopub.status.idle": "2022-08-21T11:15:56.336655Z",
          "shell.execute_reply": "2022-08-21T11:15:56.335758Z",
          "shell.execute_reply.started": "2022-08-21T11:15:56.329544Z"
        },
        "id": "_DetPu1mS2ts",
        "trusted": true
      },
      "outputs": [],
      "source": [
        "for i in enrolment_objects:\n",
        "    for j in exam_objects:\n",
        "        if j.code == i.exam_code:\n",
        "            j.student_codes.append(i.student_code)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 86,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-08-21T11:15:56.340238Z",
          "iopub.status.busy": "2022-08-21T11:15:56.338075Z",
          "iopub.status.idle": "2022-08-21T11:15:56.347590Z",
          "shell.execute_reply": "2022-08-21T11:15:56.346579Z",
          "shell.execute_reply.started": "2022-08-21T11:15:56.340210Z"
        },
        "id": "u3c4g6zrS2ts",
        "trusted": true
      },
      "outputs": [],
      "source": [
        "for i in exam_objects:\n",
        "    for j in student_objects:\n",
        "        for k in i.student_codes:\n",
        "            if k == j.code:\n",
        "                i.students.append(j)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 87,
      "metadata": {
        "id": "y0xrzX52S2tt"
      },
      "outputs": [],
      "source": [
        "for i in exam_objects:\n",
        "    for j in i.students:\n",
        "        if j.accessibility_need == True:\n",
        "            i.access_requirement = True\n",
        "            i.students_with_access_needs.append(j)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 88,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-08-21T11:15:56.419625Z",
          "iopub.status.busy": "2022-08-21T11:15:56.418914Z",
          "iopub.status.idle": "2022-08-21T11:15:56.424602Z",
          "shell.execute_reply": "2022-08-21T11:15:56.423510Z",
          "shell.execute_reply.started": "2022-08-21T11:15:56.419597Z"
        },
        "id": "BwbsEpjPS2tt",
        "trusted": true
      },
      "outputs": [],
      "source": [
        "def getRoomKey(obj):\n",
        "    return obj.capacity"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 89,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-08-21T11:15:56.505580Z",
          "iopub.status.busy": "2022-08-21T11:15:56.505011Z",
          "iopub.status.idle": "2022-08-21T11:15:56.511382Z",
          "shell.execute_reply": "2022-08-21T11:15:56.510522Z",
          "shell.execute_reply.started": "2022-08-21T11:15:56.505543Z"
        },
        "id": "Q1kWxt4CS2tu",
        "trusted": true
      },
      "outputs": [],
      "source": [
        "def hhToSeconds(s):\n",
        "    t = 0\n",
        "    splitTime = s.split(':')\n",
        "    t = 60 * int(splitTime[0]) + int(splitTime[1])\n",
        "    return t"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 90,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-08-21T11:15:56.521027Z",
          "iopub.status.busy": "2022-08-21T11:15:56.520774Z",
          "iopub.status.idle": "2022-08-21T11:15:56.525661Z",
          "shell.execute_reply": "2022-08-21T11:15:56.524671Z",
          "shell.execute_reply.started": "2022-08-21T11:15:56.521003Z"
        },
        "id": "5YlKqOMvS2tu",
        "trusted": true
      },
      "outputs": [],
      "source": [
        "for i in exam_objects:\n",
        "    i.duration_min = hhToSeconds(str(i.duration))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 91,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-08-21T11:15:56.541041Z",
          "iopub.status.busy": "2022-08-21T11:15:56.540376Z",
          "iopub.status.idle": "2022-08-21T11:15:56.547739Z",
          "shell.execute_reply": "2022-08-21T11:15:56.546792Z",
          "shell.execute_reply.started": "2022-08-21T11:15:56.541012Z"
        },
        "id": "8f9X90-gS2tu",
        "trusted": true
      },
      "outputs": [],
      "source": [
        "def degreeOfConflict(exam):\n",
        "    s = set()\n",
        "    for i in exam.students:\n",
        "        for j in i.exams:\n",
        "            if exam.code != j.code:\n",
        "                s.add(j)\n",
        "    exam.exam_conflicts = s\n",
        "    return len(s)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 92,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-08-21T11:15:56.578952Z",
          "iopub.status.busy": "2022-08-21T11:15:56.578412Z",
          "iopub.status.idle": "2022-08-21T11:15:56.583344Z",
          "shell.execute_reply": "2022-08-21T11:15:56.582450Z",
          "shell.execute_reply.started": "2022-08-21T11:15:56.578925Z"
        },
        "id": "6MiXvoU-S2tu",
        "trusted": true
      },
      "outputs": [],
      "source": [
        "for i in exam_objects:\n",
        "    i.conflicts = degreeOfConflict(i)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 93,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-08-21T11:15:56.599177Z",
          "iopub.status.busy": "2022-08-21T11:15:56.598419Z",
          "iopub.status.idle": "2022-08-21T11:15:56.609290Z",
          "shell.execute_reply": "2022-08-21T11:15:56.608395Z",
          "shell.execute_reply.started": "2022-08-21T11:15:56.599141Z"
        },
        "id": "8I8PAW85S2tv",
        "trusted": true
      },
      "outputs": [],
      "source": [
        "def partialExamOrdering():\n",
        "    sup = np.zeros(len(exam_objects))\n",
        "    for i in range(len(exam_objects)):\n",
        "        for j in range(len(exam_objects)):\n",
        "            if exam_objects[i].conflicts > exam_objects[j].conflicts:\n",
        "                sup[i] = sup[i] + 1\n",
        "                continue\n",
        "            elif exam_objects[i].conflicts < exam_objects[j].conflicts:\n",
        "                sup[j] = sup[j] + 1\n",
        "                continue\n",
        "            if len(exam_objects[i].students) > len(exam_objects[j].students):\n",
        "                sup[i] = sup[i] + 1\n",
        "                continue\n",
        "            elif len(exam_objects[i].students) < len(exam_objects[j].students):\n",
        "                sup[j] = sup[j] + 1\n",
        "                continue\n",
        "            if exam_objects[i].duration_min > exam_objects[j].duration_min:\n",
        "                sup[i] = sup[i] + 1\n",
        "                continue\n",
        "            elif exam_objects[i].duration_min < exam_objects[j].duration_min:\n",
        "                sup[j] = sup[j] + 1\n",
        "                continue\n",
        "            if len(exam_objects[i].students_with_access_needs) > len(exam_objects[j].students_with_access_needs):\n",
        "                sup[i] = sup[i] + 1\n",
        "                continue\n",
        "            elif len(exam_objects[i].students_with_access_needs) < len(exam_objects[j].students_with_access_needs):\n",
        "                sup[j] = sup[j] + 1\n",
        "                continue\n",
        "            if exam_objects[i].code > exam_objects[j].code:\n",
        "                sup[j] = sup[j] + 1\n",
        "                continue\n",
        "            elif exam_objects[i].code < exam_objects[j].code:\n",
        "                sup[i] = sup[i] + 1\n",
        "                continue\n",
        "    return sup                               "
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 94,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-08-21T11:15:56.619740Z",
          "iopub.status.busy": "2022-08-21T11:15:56.619476Z",
          "iopub.status.idle": "2022-08-21T11:15:56.624988Z",
          "shell.execute_reply": "2022-08-21T11:15:56.623771Z",
          "shell.execute_reply.started": "2022-08-21T11:15:56.619715Z"
        },
        "id": "3y-xIHkqS2tv",
        "trusted": true
      },
      "outputs": [],
      "source": [
        "def totalConflictsForExams():\n",
        "    partialExamOrder = partialExamOrdering()\n",
        "    for i in range(len(exam_objects)):\n",
        "        exam_objects[i].priority = partialExamOrder[i]  "
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 95,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-08-21T11:15:56.649514Z",
          "iopub.status.busy": "2022-08-21T11:15:56.647492Z",
          "iopub.status.idle": "2022-08-21T11:15:56.653777Z",
          "shell.execute_reply": "2022-08-21T11:15:56.652648Z",
          "shell.execute_reply.started": "2022-08-21T11:15:56.649487Z"
        },
        "id": "mLdTNsSLS2tv",
        "trusted": true
      },
      "outputs": [],
      "source": [
        "def getKey(obj):\n",
        "    return obj.priority"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 96,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-08-21T11:15:56.690465Z",
          "iopub.status.busy": "2022-08-21T11:15:56.689631Z",
          "iopub.status.idle": "2022-08-21T11:15:56.698989Z",
          "shell.execute_reply": "2022-08-21T11:15:56.698132Z",
          "shell.execute_reply.started": "2022-08-21T11:15:56.690429Z"
        },
        "id": "MDZvx42xS2tv",
        "trusted": true
      },
      "outputs": [],
      "source": [
        "pheromoneMatrix = np.zeros((len(exam_objects), len(slot_objects), len(room_objects)))\n",
        "\n",
        "def updatePheromoneMatrix(t_min, t_max, EVAPORATION_RATE):\n",
        "    for i in range(len(exam_objects)):\n",
        "        for j in range(len(slot_objects)):\n",
        "            for k in range(len(room_objects)):\n",
        "                if exam_objects[i].slot.index == slot_objects[j].index and room_objects[k] in exam_objects[i].rooms:\n",
        "                    pheromoneMatrix[i][j][k] = (1 - EVAPORATION_RATE) * pheromoneMatrix[i][j][k] + 1\n",
        "                else:\n",
        "                    pheromoneMatrix[i][j][k] = (1 - EVAPORATION_RATE) * pheromoneMatrix[i][j][k]\n",
        "                if pheromoneMatrix[i][j][k] < t_min:\n",
        "                    pheromoneMatrix[i][j][k] = t_min\n",
        "                elif pheromoneMatrix[i][j][k] > t_max:\n",
        "                    pheromoneMatrix[i][j][k] = t_max\n",
        "    \n",
        "def initPheromoneMatrix(exams, slots, rooms, value):\n",
        "    for i in range(len(exams)):\n",
        "        for j in range(len(slots)):\n",
        "            for k in range(len(rooms)):\n",
        "                pheromoneMatrix[i][j][k] = value\n",
        "       "
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 97,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-08-21T11:15:56.714814Z",
          "iopub.status.busy": "2022-08-21T11:15:56.714062Z",
          "iopub.status.idle": "2022-08-21T11:15:56.721370Z",
          "shell.execute_reply": "2022-08-21T11:15:56.720455Z",
          "shell.execute_reply.started": "2022-08-21T11:15:56.714780Z"
        },
        "id": "mi6bkz3LS2tw",
        "trusted": true
      },
      "outputs": [],
      "source": [
        "heuristicsMatrix = np.zeros((len(exam_objects), len(slot_objects), len(room_objects)))\n",
        "\n",
        "def initHeuristicsMatrix(exams, slots, rooms):\n",
        "    for i in range(len(exams)):\n",
        "        for j in range(len(slots)):\n",
        "            for k in range(len(rooms)):\n",
        "                heuristicsMatrix[i][j][k] = 1\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 98,
      "metadata": {
        "id": "DVgD1JUfS2tw"
      },
      "outputs": [],
      "source": [
        "roomCapacityMatrix = np.zeros((len(slot_objects), len(room_objects)))\n",
        "\n",
        "def initRoomCapacityMatrixMatrix(exams, slots, rooms):\n",
        "    for j in range(len(slots)):\n",
        "        for k in range(len(rooms)):\n",
        "            roomCapacityMatrix[j][k] = room_objects[k].capacity\n",
        "                \n",
        "                    "
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 99,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-08-21T11:15:56.759634Z",
          "iopub.status.busy": "2022-08-21T11:15:56.759379Z",
          "iopub.status.idle": "2022-08-21T11:15:56.765511Z",
          "shell.execute_reply": "2022-08-21T11:15:56.764423Z",
          "shell.execute_reply.started": "2022-08-21T11:15:56.759610Z"
        },
        "id": "9W6XFYmQS2tw",
        "trusted": true
      },
      "outputs": [],
      "source": [
        "def checkForHardConstraintViolations(exam, slot, room):\n",
        "    score = 0\n",
        "    for i in exam.exam_conflicts:\n",
        "            if i.slot != -1:\n",
        "                if i.slot.index == slot.index:\n",
        "                    score += 4\n",
        "                elif np.abs(i.slot.index - slot.index) < 2:\n",
        "                    score += 4\n",
        "    if slot.day == \"Saturday\" and slot.type == \"Evening\":\n",
        "        score += 4\n",
        "    if slot.type == \"Evening\" and exam.duration_min > 90:\n",
        "        score += 4\n",
        "    if len(exam.students) + slot.student_count > 2653:\n",
        "        score += 4\n",
        "    if len(exam.students) > room.capacity:\n",
        "        score += 4\n",
        "    if exam.access_requirement and room.accessible == False:\n",
        "        score += 4\n",
        "    return score\n",
        "                    "
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 100,
      "metadata": {
        "id": "mn50HTGeS2tx"
      },
      "outputs": [],
      "source": [
        "def numIncentive(exam, slot, room):\n",
        "    score = 0\n",
        "    if slot.type == \"Evening\" and exam.duration_min <= 90:\n",
        "        score += 4\n",
        "    return 0"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 101,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-08-21T11:15:56.780021Z",
          "iopub.status.busy": "2022-08-21T11:15:56.779150Z",
          "iopub.status.idle": "2022-08-21T11:15:56.785758Z",
          "shell.execute_reply": "2022-08-21T11:15:56.784831Z",
          "shell.execute_reply.started": "2022-08-21T11:15:56.779986Z"
        },
        "id": "X-1NZux0S2tx",
        "trusted": true
      },
      "outputs": [],
      "source": [
        "def checkForSoftConstraintViolations(exam, slot, room):\n",
        "    score = 0\n",
        "    for i in exam.exam_conflicts:\n",
        "        if i.slot != -1:\n",
        "            if np.abs(i.slot.index - slot.index) < 3:\n",
        "                score += 1\n",
        "    return score"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 102,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-08-21T11:15:56.806028Z",
          "iopub.status.busy": "2022-08-21T11:15:56.805349Z",
          "iopub.status.idle": "2022-08-21T11:15:56.812233Z",
          "shell.execute_reply": "2022-08-21T11:15:56.811320Z",
          "shell.execute_reply.started": "2022-08-21T11:15:56.805991Z"
        },
        "id": "lGlL-YfbS2tx",
        "trusted": true
      },
      "outputs": [],
      "source": [
        "def findForbiddenSlots(exam):\n",
        "    forbiddenSlots = set()\n",
        "    forbiddenSlots.add(slot_objects[11])\n",
        "    for i in exam.exam_conflicts:\n",
        "        if i.slot != -1 and i.code != exam.code:\n",
        "            forbiddenSlots.add(i.slot)\n",
        "    return forbiddenSlots"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 103,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-08-21T11:15:56.832895Z",
          "iopub.status.busy": "2022-08-21T11:15:56.830880Z",
          "iopub.status.idle": "2022-08-21T11:15:56.837133Z",
          "shell.execute_reply": "2022-08-21T11:15:56.836205Z",
          "shell.execute_reply.started": "2022-08-21T11:15:56.832867Z"
        },
        "id": "e2hFHYecS2ty",
        "trusted": true
      },
      "outputs": [],
      "source": [
        "def numAdditionalViolations(exam, slot, room):\n",
        "    score = 0\n",
        "    score += checkForHardConstraintViolations(exam, slot, room)\n",
        "    score += checkForSoftConstraintViolations(exam, slot, room)\n",
        "    return score"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 104,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-08-21T11:15:56.855512Z",
          "iopub.status.busy": "2022-08-21T11:15:56.854979Z",
          "iopub.status.idle": "2022-08-21T11:15:56.861314Z",
          "shell.execute_reply": "2022-08-21T11:15:56.860474Z",
          "shell.execute_reply.started": "2022-08-21T11:15:56.855476Z"
        },
        "id": "6Tyfy-5PS2ty",
        "trusted": true
      },
      "outputs": [],
      "source": [
        "def pheroHeuristicSummation(examIndex, slotIndex, roomIndex, alpha, beta):\n",
        "    sum = 0\n",
        "    for j in range(len(slot_objects)):\n",
        "        for k in range(len(room_objects)):\n",
        "            if j != slotIndex and k != roomIndex:\n",
        "                sum += (pheromoneMatrix[examIndex][j][k]**alpha) * (heuristicsMatrix[examIndex][j][k]**beta)\n",
        "    return sum"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 105,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-08-21T11:15:56.912448Z",
          "iopub.status.busy": "2022-08-21T11:15:56.911854Z",
          "iopub.status.idle": "2022-08-21T11:15:56.921353Z",
          "shell.execute_reply": "2022-08-21T11:15:56.920301Z",
          "shell.execute_reply.started": "2022-08-21T11:15:56.912405Z"
        },
        "id": "MQ_EeF2vS2ty",
        "trusted": true
      },
      "outputs": [],
      "source": [
        "def chooseSlot(examIndex, exam, slots, rooms, alpha, beta):\n",
        "    slot = 0\n",
        "    room = []\n",
        "    desirability = np.zeros((len(slots), len(rooms)))\n",
        "    forbiddenSlots = list(findForbiddenSlots(exam))\n",
        "    for i in range(len(slots)):\n",
        "        for k in range(len(rooms)):\n",
        "            if roomCapacityMatrix[i][k] <= 0:\n",
        "                heuristicsMatrix[examIndex][i][k] = 0.000000000000000000001\n",
        "            else:\n",
        "                heuristicsMatrix[examIndex][i][k] = (1 + numIncentive(exam, slots[i], rooms[k])) / (1 + numAdditionalViolations(exam, slots[i], rooms[k]))\n",
        "            desirability[i][k] = ((pheromoneMatrix[examIndex][i][k]**alpha) * (heuristicsMatrix[examIndex][i][k]**beta))\n",
        "            for j in forbiddenSlots:\n",
        "                if slot_objects[i].index == j.index:\n",
        "                    desirability[i][k] = -1\n",
        "                    continue\n",
        "    argSortedList = np.flip(np.dstack(np.unravel_index(np.argsort(desirability.ravel()), desirability.shape))[0])\n",
        "    slot = argSortedList[0][1]\n",
        "    r = argSortedList[0][0]\n",
        "    val = desirability[slot][r]\n",
        "    availableSlots = []\n",
        "    for i in argSortedList:\n",
        "        if desirability[i[1]][i[0]] == desirability[slot][r]:\n",
        "            availableSlots.append(i[1])\n",
        "    rand = np.random.randint(0, len(availableSlots))\n",
        "    slot = availableSlots[rand]\n",
        "    for index in argSortedList:\n",
        "        if len(exam.students) - exam.students_assigned == 0:\n",
        "            break\n",
        "        if index[1] == slot:\n",
        "            room_ind = index[0]\n",
        "            index_room = rooms[room_ind]\n",
        "            room.append(index_room)\n",
        "            cap = roomCapacityMatrix[slot][room_ind]\n",
        "            if len(exam.students) - exam.students_assigned >= cap:\n",
        "                exam.students_assigned += cap\n",
        "                np.asmatrix(roomCapacityMatrix)\n",
        "                roomCapacityMatrix[slot, room_ind] = 0\n",
        "            elif len(exam.students) - exam.students_assigned < cap:\n",
        "                exam.students_assigned += len(exam.students) - exam.students_assigned\n",
        "                np.asmatrix(roomCapacityMatrix)\n",
        "                roomCapacityMatrix[slot, room_ind] -= exam.students_assigned\n",
        "    return slots[slot], room"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 106,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-08-21T11:15:56.933688Z",
          "iopub.status.busy": "2022-08-21T11:15:56.933118Z",
          "iopub.status.idle": "2022-08-21T11:15:56.940801Z",
          "shell.execute_reply": "2022-08-21T11:15:56.939730Z",
          "shell.execute_reply.started": "2022-08-21T11:15:56.933651Z"
        },
        "id": "igeIgRBXS2ty",
        "trusted": true
      },
      "outputs": [],
      "source": [
        "def evaluateHardConstraints():\n",
        "    countHard = 0\n",
        "    examsChecked = []\n",
        "    index = 0\n",
        "    for i in slot_objects:\n",
        "        if i.student_count > 2653:\n",
        "            countHard += 1\n",
        "    for i in room_objects:\n",
        "        if i.seats_occupied > i.capacity:\n",
        "            countHard += 1\n",
        "    while index < len(exam_objects):\n",
        "        exam = exam_objects[index]\n",
        "        flag = False\n",
        "        for j in exam.exam_conflicts:\n",
        "            for k in examsChecked:\n",
        "                if k == j:\n",
        "                    flag = True\n",
        "            if flag:\n",
        "                continue\n",
        "            else:\n",
        "                if exam.slot.index == j.slot.index:\n",
        "                        countHard += 1\n",
        "                elif np.abs(exam.slot.index - j.slot.index) < 2:\n",
        "                    countHard += 1\n",
        "        if exam.slot.day == \"Saturday\" and exam.slot.type == \"Evening\":\n",
        "            countHard += 1\n",
        "        if exam.slot.type == \"Evening\" and exam.duration_min > 90:\n",
        "            countHard += 1\n",
        "        if exam.access_requirement:\n",
        "            for i in list(exam.rooms)[0]:\n",
        "                if i.accessible == False:\n",
        "                    countHard += 1\n",
        "        index += 1\n",
        "        examsChecked.append(exam)\n",
        "    return countHard"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 107,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-08-21T11:15:56.971491Z",
          "iopub.status.busy": "2022-08-21T11:15:56.970008Z",
          "iopub.status.idle": "2022-08-21T11:15:56.976775Z",
          "shell.execute_reply": "2022-08-21T11:15:56.975722Z",
          "shell.execute_reply.started": "2022-08-21T11:15:56.971453Z"
        },
        "id": "12hE_K-RS2tz",
        "trusted": true
      },
      "outputs": [],
      "source": [
        "def evaluateSoftConstraints():\n",
        "    countSoft = 0\n",
        "    examsChecked = []\n",
        "    index = 0\n",
        "    while index  < len(exam_objects):\n",
        "        exam = exam_objects[index]\n",
        "        flag = False\n",
        "        for j in exam.exam_conflicts:\n",
        "            for k in examsChecked:\n",
        "                if  k == j:\n",
        "                    flag = True\n",
        "            if flag:\n",
        "                continue\n",
        "            else:\n",
        "                if np.abs(exam.slot.index - j.slot.index) < 3:\n",
        "                    countSoft += 1\n",
        "        index += 1\n",
        "        examsChecked.append(exam)\n",
        "    return countSoft"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 108,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-08-21T11:15:56.982353Z",
          "iopub.status.busy": "2022-08-21T11:15:56.981429Z",
          "iopub.status.idle": "2022-08-21T11:15:56.988871Z",
          "shell.execute_reply": "2022-08-21T11:15:56.987963Z",
          "shell.execute_reply.started": "2022-08-21T11:15:56.982316Z"
        },
        "id": "8aZ6rhVSS2tz",
        "trusted": true
      },
      "outputs": [],
      "source": [
        "def evaluate():\n",
        "    metricHardConstraints = 0\n",
        "    metricSoftConstraints = 0\n",
        "    metricHardConstraints += evaluateHardConstraints()\n",
        "    metricSoftConstraints += evaluateSoftConstraints()\n",
        "    return metricSoftConstraints, metricHardConstraints"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 109,
      "metadata": {
        "id": "ZoK62EPgS2tz"
      },
      "outputs": [],
      "source": [
        "def getMapKey(obj):\n",
        "    return obj.capacity"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 110,
      "metadata": {
        "id": "t8Gc0yQ8S2tz"
      },
      "outputs": [],
      "source": [
        "room_max_capacity = room_objects[0]\n",
        "for i in room_objects:\n",
        "    if room_max_capacity.capacity < i.capacity:\n",
        "        room_max_capacity = i"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 111,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-08-21T11:15:57.265972Z",
          "iopub.status.busy": "2022-08-21T11:15:57.265642Z",
          "iopub.status.idle": "2022-08-21T11:15:57.271455Z",
          "shell.execute_reply": "2022-08-21T11:15:57.270505Z",
          "shell.execute_reply.started": "2022-08-21T11:15:57.265943Z"
        },
        "id": "jBuQJiuRS2tz",
        "trusted": true
      },
      "outputs": [],
      "source": [
        "def assignSlotsForSolution(solution):\n",
        "    for i in slot_objects:\n",
        "        i.exams = []\n",
        "        i.student_count = 0\n",
        "    for i in exam_objects:\n",
        "        i.rooms = []\n",
        "    for i in solution:\n",
        "        list(i)[0].slot = list(i)[1]\n",
        "        list(i)[0].rooms.append(list(i)[2])\n",
        "        list(i)[1].exams.append(list(i)[0])\n",
        "        list(i)[1].student_count += len(list(i)[0].students)\n",
        "        #list(i)[2].seats_occupied += len(list(i)[0].students)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 112,
      "metadata": {
        "id": "Nq_uLTKcS2t0"
      },
      "outputs": [],
      "source": [
        "t_max = 5.0\n",
        "t_min = 0.0019\n",
        "MAX_ITER = 10"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 113,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-08-21T11:15:57.289950Z",
          "iopub.status.busy": "2022-08-21T11:15:57.289637Z",
          "iopub.status.idle": "2022-08-21T11:16:22.416529Z",
          "shell.execute_reply": "2022-08-21T11:16:22.415556Z",
          "shell.execute_reply.started": "2022-08-21T11:15:57.289923Z"
        },
        "id": "EXSk3WFgS2t0",
        "trusted": true
      },
      "outputs": [],
      "source": [
        "def aco(EVAPORATION_RATE, numAnts, alpha, beta):\n",
        "    index = 0\n",
        "    totalConflictsForExams()\n",
        "    sortedExamsByPriority = sorted(exam_objects, key=getKey, reverse=True)\n",
        "    #sortedExamsByPriority = exam_objects\n",
        "    initPheromoneMatrix(sortedExamsByPriority, slot_objects, room_objects, t_max)\n",
        "    evaluationSoftMap = []\n",
        "    evaluationHardMap = []\n",
        "    A_gb = set()\n",
        "    prevScore = 0\n",
        "    while index < MAX_ITER:\n",
        "        A_ib = set()\n",
        "        prevScoreAnt = 0\n",
        "        for k in range(numAnts):\n",
        "            initHeuristicsMatrix(sortedExamsByPriority, slot_objects, room_objects)\n",
        "            initRoomCapacityMatrixMatrix(sortedExamsByPriority, slot_objects, room_objects)\n",
        "            A = set()\n",
        "            chosenSlot = 0\n",
        "            for i in range(len(exam_objects)):\n",
        "                exam_objects[i].slot = -1\n",
        "                exam_objects[i].rooms = []\n",
        "                exam_objects[i].students_assigned = 0\n",
        "            for i in slot_objects:\n",
        "                i.student_count = 0\n",
        "            for i in room_objects:\n",
        "                i.seats_occupied = 0\n",
        "                i.seats_remaining = i.capacity\n",
        "            for i in range(len(exam_objects)):\n",
        "                chosenSlot = chooseSlot(i,sortedExamsByPriority[i], slot_objects, room_objects, alpha, beta)\n",
        "                sortedExamsByPriority[i].slot = chosenSlot[0]\n",
        "                sortedExamsByPriority[i].rooms.append(chosenSlot[1])\n",
        "                chosenSlot[0].student_count += len(sortedExamsByPriority[i].students)\n",
        "                A.add((sortedExamsByPriority[i], chosenSlot[0], tuple(chosenSlot[1])))\n",
        "            assignSlotsForSolution(A)\n",
        "            scoreAl = evaluate()[0] + evaluate()[1] * 4\n",
        "            if A_ib != set():\n",
        "                A_ib = A if scoreAl < prevScoreAnt else A_ib\n",
        "                prevScoreAnt = scoreAl if scoreAl < prevScoreAnt else prevScoreAnt\n",
        "            else:\n",
        "                A_ib = A\n",
        "                prevScoreAnt = scoreAl\n",
        "        assignSlotsForSolution(A_ib)\n",
        "        scoreA_ibl = evaluate()[0] + evaluate()[1] * 4\n",
        "        if A_gb != set():\n",
        "            A_gb = A_ib if scoreA_ibl < prevScore else A_gb\n",
        "            prevScore = scoreA_ibl if scoreA_ibl < prevScore else prevScore\n",
        "        else:\n",
        "            A_gb = A_ib\n",
        "            prevScore = scoreA_ibl\n",
        "        assignSlotsForSolution(A_gb)\n",
        "        updatePheromoneMatrix(t_min, t_max, EVAPORATION_RATE)\n",
        "        evaluationSoftMap.append((index, evaluate()[0]))\n",
        "        evaluationHardMap.append((index, evaluate()[1]))\n",
        "        index += 1\n",
        "    return evaluationSoftMap, evaluationHardMap, A_gb, prevScore"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 114,
      "metadata": {
        "id": "AeKdx4WwS2t0"
      },
      "outputs": [],
      "source": [
        "aco1 = aco(0.2, 3, 1.0, 2.0)\n",
        "aco2 = aco(0.2, 3, 1.0, 4.0)\n",
        "aco3 = aco(0.2, 6, 1.0, 2.0)\n",
        "aco4 = aco(0.2, 6, 1.0, 4.0)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 129,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "gCzXnaOGS2t1",
        "outputId": "fe9033a0-67e2-48c0-8aed-a74777a7bef0"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "ACO1: 127\n",
            "ACO2: 132\n",
            "ACO3: 136\n",
            "ACO4: 125\n",
            "Best model score: 125\n"
          ]
        }
      ],
      "source": [
        "scores = [aco1,aco2,aco3,aco4]\n",
        "print(\"ACO1: \" + str(aco1[3]))\n",
        "print(\"ACO2: \" + str(aco2[3]))\n",
        "print(\"ACO3: \" + str(aco3[3]))\n",
        "print(\"ACO4: \" + str(aco4[3]))\n",
        "\n",
        "minimum = scores[0]\n",
        "for i in scores:\n",
        "    if minimum[3] > i[3]:\n",
        "        minimum = i\n",
        "\n",
        "print(\"Best model score:\", minimum[3])\n",
        "assignSlotsForSolution(minimum[2])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 116,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-08-21T11:16:22.418964Z",
          "iopub.status.busy": "2022-08-21T11:16:22.418591Z",
          "iopub.status.idle": "2022-08-21T11:16:22.627263Z",
          "shell.execute_reply": "2022-08-21T11:16:22.626212Z",
          "shell.execute_reply.started": "2022-08-21T11:16:22.418927Z"
        },
        "id": "zgBs2SicS2t2",
        "trusted": true
      },
      "outputs": [],
      "source": [
        "hard_aco1 = pd.DataFrame(aco1[1], columns=['Iterations', 'Violations'])\n",
        "hard_aco2 = pd.DataFrame(aco2[1], columns=['Iterations', 'Violations'])\n",
        "hard_aco3 = pd.DataFrame(aco3[1], columns=['Iterations', 'Violations'])\n",
        "hard_aco4 = pd.DataFrame(aco4[1], columns=['Iterations', 'Violations'])\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 117,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 479
        },
        "id": "vFpCrOLYS2t2",
        "outputId": "b8a48fa0-c6b5-49cb-8c63-a6a41cfddd8b"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x7fbf88091950>"
            ]
          },
          "execution_count": 117,
          "metadata": {},
          "output_type": "execute_result"
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 1152x576 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        }
      ],
      "source": [
        "ax = hard_aco1.plot(x='Iterations', y='Violations', figsize=(16, 8), label=\"ACO1\")\n",
        "hard_aco2.plot(ax=ax, x='Iterations', y='Violations', label=\"ACO2\")\n",
        "hard_aco3.plot(ax=ax, x='Iterations', y='Violations', label=\"ACO3\")\n",
        "hard_aco4.plot(ax=ax, x='Iterations', y='Violations', label=\"ACO4\")\n",
        "ax.legend(loc=\"upper right\")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 118,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-08-21T11:16:22.629229Z",
          "iopub.status.busy": "2022-08-21T11:16:22.628887Z",
          "iopub.status.idle": "2022-08-21T11:16:22.842846Z",
          "shell.execute_reply": "2022-08-21T11:16:22.841958Z",
          "shell.execute_reply.started": "2022-08-21T11:16:22.629185Z"
        },
        "id": "R5krmzKaS2t2",
        "trusted": true
      },
      "outputs": [],
      "source": [
        "soft_aco1 = pd.DataFrame(aco1[0], columns=['Iterations', 'Violations'])\n",
        "soft_aco2 = pd.DataFrame(aco2[0], columns=['Iterations', 'Violations'])\n",
        "soft_aco3 = pd.DataFrame(aco3[0], columns=['Iterations', 'Violations'])\n",
        "soft_aco4 = pd.DataFrame(aco4[0], columns=['Iterations', 'Violations'])\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 119,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 534
        },
        "id": "ZoHMwhwWS2t2",
        "outputId": "a551dba7-bef1-41e0-b0b4-03c74402ba43"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x7fbf881a9350>"
            ]
          },
          "execution_count": 119,
          "metadata": {},
          "output_type": "execute_result"
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 1152x576 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        }
      ],
      "source": [
        "ax = soft_aco1.plot(x='Iterations', y='Violations', figsize=(16, 8), label=\"ACO1\")\n",
        "soft_aco2.plot(ax=ax, x='Iterations', y='Violations', label=\"ACO2\")\n",
        "soft_aco3.plot(ax=ax, x='Iterations', y='Violations', label=\"ACO3\")\n",
        "soft_aco4.plot(ax=ax, x='Iterations', y='Violations', label=\"ACO4\")\n",
        "ax.legend(loc=\"upper right\")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 130,
      "metadata": {
        "id": "Wh4iiV7FS2t2"
      },
      "outputs": [],
      "source": [
        "studentArr = []\n",
        "for i in student_objects:\n",
        "    for j in i.exams:\n",
        "        roomArr = []\n",
        "        capArr = []\n",
        "        for k in list(j.rooms)[0]:\n",
        "            roomArr.append(k.code)\n",
        "            capArr.append(k.capacity)\n",
        "        studentArr.append([i.code, j.code, j.slot.index, j.slot.day, j.slot.type, j.duration, len(j.students), roomArr,capArr])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 131,
      "metadata": {
        "id": "XztlhcuKS2t3"
      },
      "outputs": [],
      "source": [
        "pd.DataFrame(studentArr, columns=[\"student_code\", \"exam_code\", \"slot_index\", \"day\", \"type\", \"duration\", \"total_students\", \"rooms\", \"rooms_capacity\" ]).to_csv('output_log/student_output_7.csv')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 132,
      "metadata": {
        "id": "xCsVl9WKS2t3"
      },
      "outputs": [],
      "source": [
        "examArr = []\n",
        "for i in exam_objects:\n",
        "    roomArr = []\n",
        "    capArr = []\n",
        "    for j in list(i.rooms)[0]:\n",
        "        roomArr.append(j.code)\n",
        "        capArr.append(j.capacity)\n",
        "    examArr.append([i.code, i.slot.index, i.slot.day, i.slot.type, i.duration, len(i.students), roomArr, capArr, i.access_requirement])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 133,
      "metadata": {
        "id": "ztzecdNgS2t3"
      },
      "outputs": [],
      "source": [
        "pd.DataFrame(examArr, columns=[\"exam_code\", \"slot_index\", \"day\", \"type\", \"duration\", \"total_students\", \"rooms\", \"rooms_capacity\", \"special_requirement\" ]).to_csv('output_log/exam_output_7.csv')"
      ]
    }
  ],
  "metadata": {
    "colab": {
      "collapsed_sections": [],
      "provenance": []
    },
    "kernelspec": {
      "display_name": "Python 3.10.6 64-bit",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.10.6"
    },
    "vscode": {
      "interpreter": {
        "hash": "95cdb06e919eab5e2c554174537356ac9b55200d1eb6f880dc25de04343a18ea"
      }
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}