Beispiel #1
0
def waitCloseDataSet(samples, robotHeight, dataset):
    for i in range(samples):
        x = 0
        y = 1
        z = uniform(1.5,2) #height
        
        path = []
            
        
        for k in range(10):
            path.append([x,y,z])
        
        trans = GeneratePath.position_transform(path,robotHeight)
        
        dataset.newSequence()
        
        for k in range(len(trans)):

            if k < 2:
                smile = 0.25
                eyebrow = 0
            elif k < 6:
                smile = 0.25 + k/10
                eyebrow = -k/8
            else:
                smile = 0.75
                eyebrow = -6/8
            
            pupils = pupilTracking(trans[k][1], trans[k][2])
            
            dataset.appendLinked(trans[k],[smile,pupils[0], pupils[1],eyebrow])
    
    return dataset
Beispiel #2
0
def leaveDataSet(samples, enviroment,robotHeight, prefSpeed, distPreferences, dataset ):
    
    xMax,xMin,yMax,yMin = enviroment
    careDist,goodDist = distPreferences
    
    for i in range(samples):
        
        z = uniform(1.5,2) #height
        
        endx = uniform(xMin,xMax)
        endy = yMax
        
        dist = uniform(goodDist,careDist)   
                
        startx = uniform(-dist,dist)
        starty = math.sqrt(dist**2 - startx**2)        
        
        path = GeneratePath.pathSpeed([startx,starty,z],[endx,endy,z],prefSpeed)

        trans = GeneratePath.position_transform(path,robotHeight)
        
        dataset.newSequence()
        
        for k in range(len(trans)):
            target = []
            
            if k  < 2:
                target.append(0.25) #smile
            else:
                target.append(-0.5)
            
            target.extend(pupilTracking(trans[k][1], trans[k][2]))
            target.append(0) #eyebrows

            dataset.appendLinked(trans[k],target)
        
    return dataset
Beispiel #3
0
def approachDataSet(samples, enviroment,robotHeight, prefSpeed, distPreferences, dataset ):
    
    xMax,xMin,yMax,yMin = enviroment
    careDist,goodDist = distPreferences
    
    for i in range(samples):
        
        z = uniform(1.5,2) #height
        
        startx = uniform(xMin,xMax)
        starty = yMax
        
        
        endx = 0
        endy = 1       
        
        path = GeneratePath.pathSpeed([startx,starty,z],[endx,endy,z],prefSpeed)

        trans = GeneratePath.position_transform(path,robotHeight)
        
        dataset.newSequence()
        
        for param in trans:
            target = []
            
            if param[0] < careDist:
                target.append(0.25) #smile
            else:
                target.append(0)
            
            target.extend(pupilTracking(param[1], param[2]))
            target.append(0) #eyebrows

            dataset.appendLinked(param,target)
        
    return dataset
Beispiel #4
0
def waitFarDataSet(samples,distPref,robotHeight,dataset):
    """
        Generates a dataset with samples number of samples, where each sample is a
        series of identical [x,y,z] cordinates chosen randomly between the "care" and
        "good" distance from the robot.
    """
    careDist,goodDist = distPref
    
    for i in range(samples):

        dist = uniform(goodDist,careDist)   
                
        x = uniform(-dist,dist)
        y = math.sqrt(dist**2 - x**2)        
        z = uniform(1.5,2) #height

        
        path = []
            
        
        for k in range(10):
            path.append([x,y,z])
        
        trans = GeneratePath.position_transform(path,robotHeight)
        
        dataset.newSequence()
        
        for k in range(len(trans)):

            if k < 2:
                smile = 0.25
                eyebrow = 0
            elif k < 6:
                smile = 0.25 - k/5
                eyebrow = k/8
            else:
                smile = -0.75
                eyebrow = 6/8
            
            pupils = pupilTracking(trans[k][1], trans[k][2])
            
            dataset.appendLinked(trans[k],[smile,pupils[0], pupils[1],eyebrow])
    
    return dataset    
        
def waitFarDataSet(samples,distPref,robotHeight,dataset):
    """
        Generates a dataset with samples number of samples, where each sample is a
        series of identical [x,y,z] cordinates chosen randomly between the "care" and
        "good" distance from the robot.
    """
    careDist,goodDist = distPref
    
    for i in range(samples):

        dist = uniform(goodDist,careDist)   
                
        x = uniform(-dist,dist)
        y = math.sqrt(dist**2 - x**2)        
        z = uniform(1.5,2) #height

        
        path = []
            
        
        for k in range(10):
            path.append([x,y,z])
        
        trans = GeneratePath.position_transform(path,robotHeight)
        
        dataset.newSequence()
        
        for k in range(len(trans)):

            if k < 2:
                smile = 0.25
                eyebrow = 0
            elif k < 6:
                smile = 0.25 - k/5
                eyebrow = k/8
            else:
                smile = -0.75
                eyebrow = 6/8
            
            pupils = pupilTracking(trans[k][1], trans[k][2])
            
            dataset.appendLinked(trans[k],[smile,pupils[0], pupils[1],eyebrow])
    
    return dataset    
Beispiel #6
0
def testFace(enviroment,tests,speedPreferences, distancePreferences, updateTime, robotHeight, network):
    
    xMax,xMin,yMax,yMin = enviroment
    careDist,goodDist = distancePreferences
    slowSpeed, prefSpeed, fastSpeed = speedPreferences
    
    random.seed()
    """
        Sets up the window for displaying the parametrized face
    """
    center = graphics.Point(200,200)
    win = graphics.GraphWin('Face', 400, 400,autoflush=False) # give title and dimensions
    win.setBackground('white')
    win.setCoords(0, 0, 400, 400)
    
    
    """
        Sets up the window for plotting the position of the person the face is looking at
    """
    fig = plt.figure()
    ax = fig.add_subplot(1, 1, 1)
    createFigure(enviroment, distancePreferences,fig,ax)    
    
    
    message = graphics.Text(graphics.Point(200, 380), 'Click to start simulation.')
    message.draw(win)
    win.getMouse()
    message.undraw()
    win.update()
    
    
    

    

    
    lingerTestPath = []
    # Linger at comfortable distance
    for i in range(tests.get("lingerFar")):
        z = random.uniform(1.5,2)
        
        startx = random.uniform(xMin,xMax)
        starty = yMax
        
        dist = random.uniform(goodDist,careDist)   
                
        endx = random.uniform(-dist,dist)
        endy = math.sqrt(dist**2 - endx**2)        
        
        
        timesLingering = random.randrange(6,15)
        
        lingerTestPath.extend(GeneratePath.moveAndLinger([startx,starty,z],[endx,endy,z],prefSpeed,timesLingering))
            
    #linger too close
    for i in range(tests.get("lingerClose")):
        z = random.uniform(1.5,2)
        startx = random.uniform(xMin,xMax)
        starty = yMax
        
        #dist = random.uniform(enviroment[3],tooClose)
       # endx = random.uniform(-dist,dist)
        #endy = math.sqrt(dist**2 - endx**2)  
        endx = 0
        endy = 1        
        timesLingering = random.randrange(6,15)
        
        lingerTestPath.extend(GeneratePath.moveAndLinger([startx,starty,z],[endx,endy,z],prefSpeed,timesLingering))
    
    

    
    trans = GeneratePath.position_transform(lingerTestPath,robotHeight)    
    
    
    print("\n---------------- Lingering tests ------------------ \n\n")
    iterateNetThroughPath(lingerTestPath, trans, prefSpeed, network, center, win, updateTime )
    
    
    
    # approach, linger, leave tests
    plt.close()

    fig = plt.figure()
    ax = fig.add_subplot(1, 1, 1)
    
    createFigure(enviroment, distancePreferences,fig,ax)
    


    approachLingerLeavePath = []
    for i in range(tests.get("approachLingerLeaveClose")):
        z = random.uniform(1.5,2)
        startx = random.uniform(xMin,xMax)
        starty = yMax

        stoppx = 0
        stoppy = 1        
        timesLingering = random.randrange(6,15)
        

        exitx = random.uniform(xMin,xMax)
        exity = yMax

        approachLingerLeavePath.extend(GeneratePath.approachLingerLeave(prefSpeed, [startx,starty,z], [stoppx,stoppy,z], [exitx,exity,z], timesLingering))


    trans = GeneratePath.position_transform(approachLingerLeavePath,robotHeight)    

    
    print("\n---------------- Approach Linger Leave Close tests ------------------ \n\n")
    iterateNetThroughPath(approachLingerLeavePath, trans, prefSpeed, network, center, win, updateTime )
    
    message.draw(win)
    win.getMouse()
    message.undraw()
    win.update()
        
    
    plt.close()

    fig = plt.figure()
    ax = fig.add_subplot(1, 1, 1)
    
    createFigure(enviroment, distancePreferences,fig,ax)




    approachLingerLeavePath = []
    
    for i in range(tests.get("approachLingerLeaveFar")):
        z = random.uniform(1.5,2)
        startx = random.uniform(xMin,xMax)
        starty = yMax
        
        dist = random.uniform(goodDist,careDist)
        stoppx = random.uniform(-dist,dist)
        stoppy = math.sqrt(dist**2 - stoppx**2)  

        timesLingering = random.randrange(6,15)
        
        exitx = random.uniform(xMin,xMax)
        exity = yMax

        approachLingerLeavePath.extend(GeneratePath.approachLingerLeave(prefSpeed,[startx,starty,z], [stoppx,stoppy,z], [exitx,exity,z], timesLingering))
    
    
    trans = GeneratePath.position_transform(approachLingerLeavePath,robotHeight)    

    
    print("\n---------------- Approach Linger Leave Far tests ------------------ \n\n")  
    iterateNetThroughPath(approachLingerLeavePath, trans, prefSpeed, network, center, win, updateTime )
    
    message.draw(win)
    win.getMouse()
    message.undraw()
    win.update()
        
    
    plt.close()

    fig = plt.figure()
    ax = fig.add_subplot(1, 1, 1)
    
    createFigure(enviroment, distancePreferences,fig,ax)

    
    #random movement tests
    path = GeneratePath.randomMovement(enviroment,tests.get("randomMove"),random.uniform(1.5,2),prefSpeed)
    trans = GeneratePath.position_transform(path,robotHeight)
    
    print("\n---------------- Random movement tests ------------------ \n\n")
    iterateNetThroughPath(path, trans, prefSpeed, network, center, win, updateTime )
    

        
        
    message = graphics.Text(graphics.Point(200, 380), 'Click to close window.')
    message.draw(win)
    win.getMouse()
    
    win.close()
    plt.close()
def perform(level, box, options):

    logging.info("BoundingBox coordinates: ({},{}),({},{}),({},{})".format(
        box.miny, box.maxy, box.minx, box.maxx, box.minz, box.maxz))

    # ==== PREPARATION =====
    (width, height, depth) = utilityFunctions.getBoxSize(box)
    logging.info("Selection box dimensions {}, {}, {}".format(
        width, height, depth))
    world = utilityFunctions.generateMatrix(level, box, width, depth, height)
    world_space = utilityFunctions.dotdict({
        "y_min": 0,
        "y_max": height - 1,
        "x_min": 0,
        "x_max": width - 1,
        "z_min": 0,
        "z_max": depth - 1
    })
    height_map = utilityFunctions.getHeightMap(level, box)

    # ==== PARTITIONING OF NEIGHBOURHOODS ====
    (center,
     neighbourhoods) = generateCenterAndNeighbourhood(world_space, height_map)
    all_buildings = []

    # ====  GENERATING CITY CENTER ====
    minimum_h = 50
    minimum_w = 25
    mininum_d = 25

    iterate = 100
    minimum_lots = 6
    available_lots = 0
    maximum_tries = 50
    current_try = 0
    threshold = 1
    partitioning_list = []
    temp_partitioning_list = []

    # run the partitioning algorithm for iterate times to get different partitionings of the same area
    logging.info(
        "Generating {} different partitionings for the the City Centre {}".
        format(iterate, center))
    while available_lots < minimum_lots and current_try < maximum_tries:

        for i in range(iterate):

            # generate a partitioning through some algorithm
            if RNG.random() < 0.5:
                partitioning = binarySpacePartitioning(center[0], center[1],
                                                       center[2], center[3],
                                                       center[4], center[5],
                                                       [])
            else:
                partitioning = quadtreeSpacePartitioning(
                    center[0], center[1], center[2], center[3], center[4],
                    center[5], [])

            # remove invalid partitions from the partitioning
            valid_partitioning = []
            for p in partitioning:
                (y_min, y_max, x_min, x_max, z_min, z_max) = (p[0], p[1], p[2],
                                                              p[3], p[4], p[5])
                failed_conditions = []
                cond1 = utilityFunctions.hasValidGroundBlocks(
                    x_min, x_max, z_min, z_max, height_map)
                if cond1 == False: failed_conditions.append(1)
                cond2 = utilityFunctions.hasMinimumSize(
                    y_min, y_max, x_min, x_max, z_min, z_max, minimum_h,
                    minimum_w, mininum_d)
                if cond2 == False: failed_conditions.append(2)
                cond3 = utilityFunctions.hasAcceptableSteepness(
                    x_min, x_max, z_min, z_max, height_map,
                    utilityFunctions.getScoreArea_type1, threshold)
                if cond3 == False: failed_conditions.append(3)
                if cond1 and cond2 and cond3:
                    score = utilityFunctions.getScoreArea_type1(
                        height_map, x_min, x_max, z_min, z_max)
                    valid_partitioning.append((score, p))
                else:
                    logging.info(
                        "Failed Conditions {}".format(failed_conditions))

            partitioning_list.extend(valid_partitioning)
            logging.info(
                "Generated a partition with {} valid lots and {} invalids ones"
                .format(len(valid_partitioning),
                        len(partitioning) - len(valid_partitioning)))

        # order partitions by steepness
        partitioning_list = sorted(partitioning_list)
        final_partitioning = utilityFunctions.getNonIntersectingPartitions(
            partitioning_list)

        available_lots = len(final_partitioning)
        logging.info(
            "Current partitioning with most available_lots: {}, current threshold {}"
            .format(available_lots, threshold))

        threshold += 1
        current_try += 1

    logging.info("Final lots ({}) for the City Centre {}: ".format(
        len(final_partitioning), center))
    for p in final_partitioning:
        logging.info("\t{}".format(p))

    for partition in final_partitioning:
        building = generateBuilding(world, partition, height_map)
        all_buildings.append(building)

    # ==== GENERATING NEIGHBOURHOODS ====

    minimum_h = 3
    minimum_w = 7
    mininum_d = 6

    iterate = 100
    maximum_tries = 50
    current_try = 0
    minimum_lots = 20
    available_lots = 0
    threshold = 1
    partitioning_list = []
    final_partitioning = []

    while available_lots < minimum_lots and current_try < maximum_tries:
        partitioning_list = []
        for i in range(iterate):
            for neigh in neighbourhoods:
                logging.info(
                    "Generating {} different partitionings for the neighbourhood {}"
                    .format(iterate, neigh))

                if RNG.random() < 0.5:
                    partitioning = binarySpacePartitioning(
                        neigh[0], neigh[1], neigh[2], neigh[3], neigh[4],
                        neigh[5], [])
                else:
                    partitioning = quadtreeSpacePartitioning(
                        neigh[0], neigh[1], neigh[2], neigh[3], neigh[4],
                        neigh[5], [])

                valid_partitioning = []
                for p in partitioning:
                    (y_min, y_max, x_min, x_max, z_min,
                     z_max) = (p[0], p[1], p[2], p[3], p[4], p[5])
                    failed_conditions = []
                    cond1 = utilityFunctions.hasValidGroundBlocks(
                        x_min, x_max, z_min, z_max, height_map)
                    if cond1 == False: failed_conditions.append(1)
                    cond2 = utilityFunctions.hasMinimumSize(
                        y_min, y_max, x_min, x_max, z_min, z_max, minimum_h,
                        minimum_w, mininum_d)
                    if cond2 == False: failed_conditions.append(2)
                    cond3 = utilityFunctions.hasAcceptableSteepness(
                        x_min, x_max, z_min, z_max, height_map,
                        utilityFunctions.getScoreArea_type1, threshold)
                    if cond3 == False: failed_conditions.append(3)
                    if cond1 and cond2 and cond3:
                        score = utilityFunctions.getScoreArea_type1(
                            height_map, x_min, x_max, z_min, z_max)
                        valid_partitioning.append((score, p))
                        logging.info("Passed the 3 conditions!")
                    else:
                        logging.info(
                            "Failed Conditions {}".format(failed_conditions))

                partitioning_list.extend(valid_partitioning)
                logging.info(
                    "Generated a partition with {} valid lots and {} invalids ones"
                    .format(len(valid_partitioning),
                            len(partitioning) - len(valid_partitioning)))

        temp_partitioning_list.extend(partitioning_list)
        # order partitions by steepness
        temp_partitioning_list = sorted(temp_partitioning_list)
        final_partitioning = utilityFunctions.getNonIntersectingPartitions(
            temp_partitioning_list)

        available_lots = len(final_partitioning)
        logging.info(
            "Current neighbourhood partitioning with most available_lots: {}, current threshold {}"
            .format(available_lots, threshold))

        threshold += 1
        current_try += 1

        logging.info("Final lots ({})for the neighbourhood {}: ".format(
            len(final_partitioning), neigh))
        for p in final_partitioning:
            logging.info("\t{}".format(p))

    for partition in final_partitioning:
        house = generateHouse(world, partition, height_map)
        all_buildings.append(house)

    # ==== GENERATE PATH MAP  ====
    # generate a path map that gives the cost of moving to each neighbouring cell
    pathMap = utilityFunctions.getPathMap(height_map, width, depth)

    # ==== CONNECTING BUILDINGS WITH ROADS  ====
    logging.info("Calling MST on {} buildings".format(len(all_buildings)))
    MST = utilityFunctions.getMST_Manhattan(all_buildings, pathMap, height_map)

    pavementBlockID = 4
    pavementBlockSubtype = 0
    for m in MST:
        p1 = m[1]
        p2 = m[2]
        logging.info("Pavement with distance {} between {} and {}".format(
            m[0], p1.entranceLot, p2.entranceLot))

        path = utilityFunctions.aStar(p1.entranceLot, p2.entranceLot, pathMap,
                                      height_map)
        if path != None:
            logging.info(
                "Found path between {} and {}. Generating road...".format(
                    p1.entranceLot, p2.entranceLot))
            GeneratePath.generatPath(world, path, height_map,
                                     (pavementBlockID, pavementBlockSubtype))
        else:
            logging.info(
                "Couldnt find path between {} and {}. Generating a straight road between them..."
                .format(p1.entranceLot, p2.entranceLot))
            GeneratePath.generatPath_StraightLine(
                world, p1.entranceLot[1], p1.entranceLot[2], p2.entranceLot[1],
                p2.entranceLot[2], height_map,
                (pavementBlockID, pavementBlockSubtype))

    # ==== UPDATE WORLD ====
    world.updateWorld()
Beispiel #8
0
def testFace(enviroment, tests, speedPreferences, distancePreferences,
             updateTime, robotHeight, network):

    xMax, xMin, yMax, yMin = enviroment
    careDist, goodDist = distancePreferences
    slowSpeed, prefSpeed, fastSpeed = speedPreferences

    random.seed()
    """
        Sets up the window for displaying the parametrized face
    """
    center = graphics.Point(200, 200)
    win = graphics.GraphWin('Face', 400, 400,
                            autoflush=False)  # give title and dimensions
    win.setBackground('white')
    win.setCoords(0, 0, 400, 400)
    """
        Sets up the window for plotting the position of the person the face is looking at
    """
    fig = plt.figure()
    ax = fig.add_subplot(1, 1, 1)
    createFigure(enviroment, distancePreferences, fig, ax)

    message = graphics.Text(graphics.Point(200, 380),
                            'Click to start simulation.')
    message.draw(win)
    win.getMouse()
    message.undraw()
    win.update()

    lingerTestPath = []
    # Linger at comfortable distance
    for i in range(tests.get("lingerFar")):
        z = random.uniform(1.5, 2)

        startx = random.uniform(xMin, xMax)
        starty = yMax

        dist = random.uniform(goodDist, careDist)

        endx = random.uniform(-dist, dist)
        endy = math.sqrt(dist**2 - endx**2)

        timesLingering = random.randrange(6, 15)

        lingerTestPath.extend(
            GeneratePath.moveAndLinger([startx, starty, z], [endx, endy, z],
                                       prefSpeed, timesLingering))

    #linger too close
    for i in range(tests.get("lingerClose")):
        z = random.uniform(1.5, 2)
        startx = random.uniform(xMin, xMax)
        starty = yMax

        #dist = random.uniform(enviroment[3],tooClose)
        # endx = random.uniform(-dist,dist)
        #endy = math.sqrt(dist**2 - endx**2)
        endx = 0
        endy = 1
        timesLingering = random.randrange(6, 15)

        lingerTestPath.extend(
            GeneratePath.moveAndLinger([startx, starty, z], [endx, endy, z],
                                       prefSpeed, timesLingering))

    trans = GeneratePath.position_transform(lingerTestPath, robotHeight)

    print("\n---------------- Lingering tests ------------------ \n\n")
    iterateNetThroughPath(lingerTestPath, trans, prefSpeed, network, center,
                          win, updateTime)

    # approach, linger, leave tests
    plt.close()

    fig = plt.figure()
    ax = fig.add_subplot(1, 1, 1)

    createFigure(enviroment, distancePreferences, fig, ax)

    approachLingerLeavePath = []
    for i in range(tests.get("approachLingerLeaveClose")):
        z = random.uniform(1.5, 2)
        startx = random.uniform(xMin, xMax)
        starty = yMax

        stoppx = 0
        stoppy = 1
        timesLingering = random.randrange(6, 15)

        exitx = random.uniform(xMin, xMax)
        exity = yMax

        approachLingerLeavePath.extend(
            GeneratePath.approachLingerLeave(prefSpeed, [startx, starty, z],
                                             [stoppx, stoppy, z],
                                             [exitx, exity, z],
                                             timesLingering))

    trans = GeneratePath.position_transform(approachLingerLeavePath,
                                            robotHeight)

    print(
        "\n---------------- Approach Linger Leave Close tests ------------------ \n\n"
    )
    iterateNetThroughPath(approachLingerLeavePath, trans, prefSpeed, network,
                          center, win, updateTime)

    message.draw(win)
    win.getMouse()
    message.undraw()
    win.update()

    plt.close()

    fig = plt.figure()
    ax = fig.add_subplot(1, 1, 1)

    createFigure(enviroment, distancePreferences, fig, ax)

    approachLingerLeavePath = []

    for i in range(tests.get("approachLingerLeaveFar")):
        z = random.uniform(1.5, 2)
        startx = random.uniform(xMin, xMax)
        starty = yMax

        dist = random.uniform(goodDist, careDist)
        stoppx = random.uniform(-dist, dist)
        stoppy = math.sqrt(dist**2 - stoppx**2)

        timesLingering = random.randrange(6, 15)

        exitx = random.uniform(xMin, xMax)
        exity = yMax

        approachLingerLeavePath.extend(
            GeneratePath.approachLingerLeave(prefSpeed, [startx, starty, z],
                                             [stoppx, stoppy, z],
                                             [exitx, exity, z],
                                             timesLingering))

    trans = GeneratePath.position_transform(approachLingerLeavePath,
                                            robotHeight)

    print(
        "\n---------------- Approach Linger Leave Far tests ------------------ \n\n"
    )
    iterateNetThroughPath(approachLingerLeavePath, trans, prefSpeed, network,
                          center, win, updateTime)

    message.draw(win)
    win.getMouse()
    message.undraw()
    win.update()

    plt.close()

    fig = plt.figure()
    ax = fig.add_subplot(1, 1, 1)

    createFigure(enviroment, distancePreferences, fig, ax)

    #random movement tests
    path = GeneratePath.randomMovement(enviroment, tests.get("randomMove"),
                                       random.uniform(1.5, 2), prefSpeed)
    trans = GeneratePath.position_transform(path, robotHeight)

    print("\n---------------- Random movement tests ------------------ \n\n")
    iterateNetThroughPath(path, trans, prefSpeed, network, center, win,
                          updateTime)

    message = graphics.Text(graphics.Point(200, 380), 'Click to close window.')
    message.draw(win)
    win.getMouse()

    win.close()
    plt.close()
def perform(level, box, options):
    logging.info("BoundingBox coordinates: ({},{}),({},{}),({},{})".format(
        box.miny, box.maxy, box.minx, box.maxx, box.minz, box.maxz))

    # ==== PREPARATION =====
    logging.info("Preparation")
    (width, height, depth) = utilityFunctions.getBoxSize(box)
    logging.info("Selection box dimensions {}, {}, {}".format(
        width, height, depth))
    world = utilityFunctions.generateMatrix(level, box, width, depth, height)
    world_space = utilityFunctions.dotdict({
        "y_min": 0,
        "y_max": height - 1,
        "x_min": 0,
        "x_max": width - 1,
        "z_min": 0,
        "z_max": depth - 1
    })
    logging.info("Generating simple height map")
    simple_height_map = utilityFunctions.getSimpleHeightMap(
        level, box)  #no height = -1 when water like block
    logging.info("Saving and erasing the trees")
    list_trees = TreeGestion.prepareMap(
        world, simple_height_map
    )  #get a list of all trees and erase them, so we can put some of them back after
    logging.info("Generating normal height map")
    height_map = utilityFunctions.getHeightMap(level, box)
    #villageDeck = utilityFunctions.generateVillageDeck("city", width, depth)

    # ==== PARTITIONING OF NEIGHBOURHOODS ====
    logging.info(
        "Partitioning of the map, getting city center and neighbourhoods")
    (center,
     neighbourhoods) = generateCenterAndNeighbourhood(world_space, height_map)
    all_buildings = []

    # ====  GENERATING CITY CENTER ====
    logging.info("Generating city center")
    minimum_h = 50
    minimum_w = 25
    mininum_d = 25

    iterate = 100
    minimum_lots = 6
    available_lots = 0
    maximum_tries = 50
    current_try = 0
    threshold = 20
    partitioning_list = []
    temp_partitioning_list = []

    # run the partitioning algorithm for iterate times to get different partitionings of the same area
    logging.info(
        "Generating {} different partitionings for the the City Centre {}".
        format(iterate, center))
    while available_lots < minimum_lots and current_try < maximum_tries:

        for i in range(iterate):

            # generate a partitioning through some algorithm
            if RNG.random() < 0.5:
                partitioning = binarySpacePartitioning(center[0], center[1],
                                                       center[2], center[3],
                                                       center[4], center[5],
                                                       [])
            else:
                partitioning = quadtreeSpacePartitioning(
                    center[0], center[1], center[2], center[3], center[4],
                    center[5], [])

            # remove invalid partitions from the partitioning
            valid_partitioning = []
            for p in partitioning:
                (y_min, y_max, x_min, x_max, z_min, z_max) = (p[0], p[1], p[2],
                                                              p[3], p[4], p[5])
                failed_conditions = []
                cond1 = utilityFunctions.hasValidGroundBlocks(
                    x_min, x_max, z_min, z_max, height_map)
                if cond1 == False: failed_conditions.append(1)
                cond2 = utilityFunctions.hasMinimumSize(
                    y_min, y_max, x_min, x_max, z_min, z_max, minimum_h,
                    minimum_w, mininum_d)
                if cond2 == False: failed_conditions.append(2)
                cond3 = utilityFunctions.hasAcceptableSteepness(
                    x_min, x_max, z_min, z_max, height_map,
                    utilityFunctions.getScoreArea_type4, threshold)
                if cond3 == False: failed_conditions.append(3)
                if cond1 and cond2 and cond3:
                    score = utilityFunctions.getScoreArea_type4(
                        height_map, x_min, x_max, z_min, z_max)
                    valid_partitioning.append((score, p))
                    logging.info("Passed the 3 conditions!")
                else:
                    logging.info(
                        "Failed Conditions {}".format(failed_conditions))

            partitioning_list.extend(valid_partitioning)
            logging.info(
                "Generated a partition with {} valid lots and {} invalids ones"
                .format(len(valid_partitioning),
                        len(partitioning) - len(valid_partitioning)))

        # order partitions by steepness
        partitioning_list = sorted(partitioning_list)
        final_partitioning = utilityFunctions.getNonIntersectingPartitions(
            partitioning_list)

        available_lots = len(final_partitioning)
        logging.info(
            "Current partitioning with most available_lots: {}, current threshold {}"
            .format(available_lots, threshold))

        threshold += 2
        current_try += 1

    logging.info("Final lots ({}) for the City Centre {}: ".format(
        len(final_partitioning), center))
    for p in final_partitioning:
        logging.info("\t{}".format(p))

    for partition in final_partitioning:
        building = generateBuilding(world, partition, height_map,
                                    simple_height_map)
        all_buildings.append(building)

    # ==== GENERATING NEIGHBOURHOODS ====
    logging.info("Generating neighbourhoods")
    minimum_h = 10
    minimum_w = 16
    mininum_d = 16

    iterate = 100
    maximum_tries = 80
    current_try = 0
    minimum_lots = 50
    available_lots = 0
    threshold = 50
    partitioning_list = []
    final_partitioning = []

    while available_lots < minimum_lots and current_try < maximum_tries:
        partitioning_list = []
        for i in range(iterate):
            for neigh in neighbourhoods:
                logging.info(
                    "Generating {} different partitionings for the neighbourhood {}"
                    .format(iterate, neigh))

                if RNG.random() < 0.5:
                    partitioning = binarySpacePartitioning(
                        neigh[0], neigh[1], neigh[2], neigh[3], neigh[4],
                        neigh[5], [])
                else:
                    partitioning = quadtreeSpacePartitioning(
                        neigh[0], neigh[1], neigh[2], neigh[3], neigh[4],
                        neigh[5], [])

                valid_partitioning = []
                for p in partitioning:
                    (y_min, y_max, x_min, x_max, z_min,
                     z_max) = (p[0], p[1], p[2], p[3], p[4], p[5])
                    failed_conditions = []
                    cond1 = utilityFunctions.hasValidGroundBlocks(
                        x_min, x_max, z_min, z_max, height_map)
                    if cond1 == False: failed_conditions.append(1)
                    cond2 = utilityFunctions.hasMinimumSize(
                        y_min, y_max, x_min, x_max, z_min, z_max, minimum_h,
                        minimum_w, mininum_d)
                    if cond2 == False: failed_conditions.append(2)
                    cond3 = utilityFunctions.hasAcceptableSteepness(
                        x_min, x_max, z_min, z_max, height_map,
                        utilityFunctions.getScoreArea_type4, threshold)
                    if cond3 == False: failed_conditions.append(3)
                    if cond1 and cond2 and cond3:
                        score = utilityFunctions.getScoreArea_type4(
                            height_map, x_min, x_max, z_min, z_max)
                        valid_partitioning.append((score, p))
                        logging.info("Passed the 3 conditions!")
                    else:
                        logging.info(
                            "Failed Conditions {}".format(failed_conditions))

                partitioning_list.extend(valid_partitioning)
                logging.info(
                    "Generated a partition with {} valid lots and {} invalids ones"
                    .format(len(valid_partitioning),
                            len(partitioning) - len(valid_partitioning)))

        temp_partitioning_list.extend(partitioning_list)
        # order partitions by steepness
        temp_partitioning_list = sorted(temp_partitioning_list)
        final_partitioning = utilityFunctions.getNonIntersectingPartitions(
            temp_partitioning_list)

        available_lots = len(final_partitioning)
        logging.info(
            "Current neighbourhood partitioning with most available_lots: {}, current threshold {}"
            .format(available_lots, threshold))

        threshold += 2
        current_try += 1

        logging.info("Final lots ({})for the neighbourhood {}: ".format(
            len(final_partitioning), neigh))
        for p in final_partitioning:
            logging.info("\t{}".format(p))

    logging.info("Building in the neighbourhood")
    n = 0
    for i in xrange(0, int(len(final_partitioning) * 0.50) + 1):
        house = generateHouse(world, final_partitioning[i], height_map,
                              simple_height_map)
        all_buildings.append(house)
        logging.info("House number : {} built on lot number {}".format(
            n + 1, i + 1))
        n += 1
    n = 0
    for i in xrange(
            int(len(final_partitioning) * 0.50) + 1,
            int(len(final_partitioning) * 0.70) + 1):
        # generate either a regular farm or a smiley farm
        farm = generateFarm(world, final_partitioning[i],
                            height_map, simple_height_map) if (RNG.randint(
                                0, 2) == 0) else generateFarm(
                                    world, final_partitioning[i], height_map,
                                    simple_height_map, "smiley")
        all_buildings.append(farm)
        logging.info("Farm number : {} built on lot number {}".format(
            n + 1, i + 1))
        n += 1
    n = 0
    m = 0
    for i in xrange(
            int(len(final_partitioning) * 0.70) + 1, len(final_partitioning)):
        slopeStructure = generateSlopeStructure(world, final_partitioning[i],
                                                height_map, simple_height_map)
        if slopeStructure.type == "tower":
            all_buildings.append(slopeStructure)
            logging.info("Tower number : {} built on lot number {}".format(
                n + 1, i + 1))
            n += 1
        else:
            logging.info(
                "RollerCoaster number : {} built on lot number {}".format(
                    m + 1, i + 1))
            m += 1

    # ==== GENERATE PATH MAP  ====
    # generate a path map that gives the cost of moving to each neighbouring cell
    logging.info("Generating path map and simple path map")
    pathMap = utilityFunctions.getPathMap(height_map, width, depth)
    simple_pathMap = utilityFunctions.getPathMap(simple_height_map, width,
                                                 depth)  #not affected by water

    # ==== CONNECTING BUILDINGS WITH ROADS  ====
    logging.info("Calling MST on {} buildings".format(len(all_buildings)))
    MST = utilityFunctions.getMST_Manhattan(all_buildings)

    for m in MST:
        p1 = m[1]
        p2 = m[2]
        if p1.type == "farm" or p2.type == "farm":
            pavement_Type = "Grass"
            bridge_Type = "Wood"
        else:
            pavement_Type = "Stone"
            bridge_Type = "Stone"

        try:
            logging.info(
                "Trying to find a path between {} and {}, finding potential bridges"
                .format(p1.entranceLot, p2.entranceLot))
            simple_path = utilityFunctions.simpleAStar(
                p1.entranceLot, p2.entranceLot, simple_pathMap,
                simple_height_map)  #water and height are not important
            list_end_points = utilityFunctions.findBridgeEndPoints(
                world, simple_path, simple_height_map)

            if list_end_points != []:
                for i in xrange(0, len(list_end_points), 2):
                    logging.info(
                        "Found water between {} and {}. Trying to generating a {} bridge..."
                        .format(list_end_points[i], list_end_points[i + 1],
                                bridge_Type))
                    GenerateBridge.generateBridge(world, simple_height_map,
                                                  list_end_points[i],
                                                  list_end_points[i + 1],
                                                  bridge_Type)
                list_end_points.insert(0, p1.entranceLot)
                list_end_points.append(p2.entranceLot)
                for i in xrange(0, len(list_end_points), 2):
                    path = utilityFunctions.aStar(list_end_points[i],
                                                  list_end_points[i + 1],
                                                  pathMap, height_map)
                    logging.info(
                        "Connecting end points of the bridge(s), Generating {} road between {} and {}"
                        .format(pavement_Type, list_end_points[i],
                                list_end_points[i + 1]))
                    GeneratePath.generatePath(world, path, height_map,
                                              pavement_Type)
            else:
                logging.info(
                    "No potential bridge found, Generating road between {} and {}"
                    .format(list_end_points[i], list_end_points[i + 1]))
                GeneratePath.generatePath(world, simple_path, height_map,
                                          pavement_Type)

        except:
            logging.info(
                "Bridge found but is not buildable, Trying to find a path between {} and {} avoiding water"
                .format(p1.entranceLot, p2.entranceLot))
            path = utilityFunctions.aStar(p1.entranceLot, p2.entranceLot,
                                          pathMap, height_map)
            if path != None:
                logging.info(
                    "Path found, Generating {} road between {} and {}".format(
                        pavement_Type, p1.entranceLot, p2.entranceLot))
                GeneratePath.generatePath(world, path, height_map,
                                          pavement_Type)
            else:
                logging.info(
                    "Couldnt find path between {} and {}. Generating a straight road"
                    .format(p1.entranceLot, p2.entranceLot))
                GeneratePath.generatePath_StraightLine(
                    world, p1.entranceLot[1], p1.entranceLot[2],
                    p2.entranceLot[1], p2.entranceLot[2], height_map,
                    pavement_Type)

    # ==== PUT BACK UNTOUCHED TREES ====
    logging.info("Putting back untouched trees")
    TreeGestion.putBackTrees(
        world, height_map, list_trees
    )  #put back the trees that are not cut buy the building and are not in unwanted places

    # ==== UPDATE WORLD ====
    world.updateWorld()
Beispiel #10
0
def perform(level, box, options):

    logging.info("BoundingBox coordinates: ({},{}),({},{}),({},{})".format(
        box.miny, box.maxy, box.minx, box.maxx, box.minz, box.maxz))

    # ==== PREPARATION =====
    (width, height, depth) = utilityFunctions.getBoxSize(box)
    logging.info("Selection box dimensions {}, {}, {}".format(
        width, height, depth))
    world = utilityFunctions.generateMatrix(level, box, width, depth, height)
    world_space = utilityFunctions.dotdict({
        "y_min": 0,
        "y_max": height - 1,
        "x_min": 0,
        "x_max": width - 1,
        "z_min": 0,
        "z_max": depth - 1
    })
    height_map = utilityFunctions.getHeightMap(level, box, None, False)

    # === Wood quantity and Biome analyzer === #
    air_like = [
        0, 4, 5, 6, 20, 23, 29, 30, 35, 37, 38, 39, 40, 44, 46, 47, 50, 59, 66,
        83, 85, 86, 95, 102, 104, 105, 107, 126, 141, 142, 160, 175
    ]
    wood_height_map = utilityFunctions.getHeightMap(level, box, air_like, True)
    (usable_wood, biome) = EnvironmentAnalyzer.determinate_usable_wood(
        level, wood_height_map, box.minx, box.maxx, box.minz, box.maxz)

    # ===  City stats === #
    biome_with_well = ['Desert', 'Badlands']
    APARTMENT_SIZE = 2
    HOUSE_SIZE = 4
    GREENHOUSE_CAPACITY = 15
    inhabitants = 0
    greenhouse_count = 0
    well_count = 0

    # ==== PARTITIONING OF THE SELECTION IN AREA IN ORDER TO FLATTEN ====
    # Work in progress do not use
    #earthwork_height_map = utilityFunctions.getHeightMap(level,box, None, True)
    #area_partitioning_and_flattening(world_space, earthwork_height_map, world, biome)

    # ==== PARTITIONING OF NEIGHBOURHOODS ====
    (center,
     neighbourhoods) = generateCenterAndNeighbourhood(world_space, height_map)
    all_buildings = []

    # ====  GENERATING CITY CENTER ====
    minimum_h = 50
    minimum_w = 25
    mininum_d = 25

    iterate = 100
    minimum_lots = 6
    available_lots = 0
    maximum_tries = 50
    current_try = 0
    threshold = 1
    partitioning_list = []
    temp_partitioning_list = []

    # run the partitioning algorithm for iterate times to get different partitionings of the same area
    logging.info(
        "Generating {} different partitionings for the the City Centre {}".
        format(iterate, center))
    while available_lots < minimum_lots and current_try < maximum_tries:

        for i in range(iterate):

            # generate a partitioning through some algorithm
            if RNG.random() < 0.5:
                partitioning = binarySpacePartitioning(center[0], center[1],
                                                       center[2], center[3],
                                                       center[4], center[5],
                                                       [])
            else:
                partitioning = quadtreeSpacePartitioning(
                    center[0], center[1], center[2], center[3], center[4],
                    center[5], [])

            # remove invalid partitions from the partitioning
            valid_partitioning = []
            for p in partitioning:
                (y_min, y_max, x_min, x_max, z_min, z_max) = (p[0], p[1], p[2],
                                                              p[3], p[4], p[5])
                failed_conditions = []
                cond1 = utilityFunctions.hasValidGroundBlocks(
                    x_min, x_max, z_min, z_max, height_map)
                if cond1 == False: failed_conditions.append(1)
                cond2 = utilityFunctions.hasMinimumSize(
                    y_min, y_max, x_min, x_max, z_min, z_max, minimum_h,
                    minimum_w, mininum_d)
                if cond2 == False: failed_conditions.append(2)
                cond3 = utilityFunctions.hasAcceptableSteepness(
                    x_min, x_max, z_min, z_max, height_map,
                    utilityFunctions.getScoreArea_type1, threshold)
                if cond3 == False: failed_conditions.append(3)
                if cond1 and cond2 and cond3:
                    score = utilityFunctions.getScoreArea_type1(
                        height_map, x_min, x_max, z_min, z_max)
                    valid_partitioning.append((score, p))
                else:
                    logging.info(
                        "Failed Conditions {}".format(failed_conditions))

            partitioning_list.extend(valid_partitioning)
            logging.info(
                "Generated a partition with {} valid lots and {} invalids ones"
                .format(len(valid_partitioning),
                        len(partitioning) - len(valid_partitioning)))

        # order partitions by steepness
        partitioning_list = sorted(partitioning_list)
        final_partitioning = utilityFunctions.getNonIntersectingPartitions(
            partitioning_list)

        available_lots = len(final_partitioning)
        logging.info(
            "Current partitioning with most available_lots: {}, current threshold {}"
            .format(available_lots, threshold))

        threshold += 1
        current_try += 1

    logging.info("Final lots ({}) for the City Centre {}: ".format(
        len(final_partitioning), center))
    for p in final_partitioning:
        logging.info("\t{}".format(p))

    for partition in final_partitioning:
        building, apartments_inhabitants = generateBuilding(
            world, partition, height_map, usable_wood, biome)
        inhabitants += apartments_inhabitants
        all_buildings.append(building)

    # ==== GENERATING NEIGHBOURHOODS ====

    minimum_h = 10
    minimum_w = 16
    mininum_d = 16

    iterate = 100
    maximum_tries = 50
    current_try = 0
    minimum_lots = 20
    available_lots = 0
    threshold = 1
    partitioning_list = []
    final_partitioning = []

    while available_lots < minimum_lots and current_try < maximum_tries:
        partitioning_list = []
        for i in range(iterate):
            for neigh in neighbourhoods:
                logging.info(
                    "Generating {} different partitionings for the neighbourhood {}"
                    .format(iterate, neigh))

                if RNG.random() < 0.5:
                    partitioning = binarySpacePartitioning(
                        neigh[0], neigh[1], neigh[2], neigh[3], neigh[4],
                        neigh[5], [])
                else:
                    partitioning = quadtreeSpacePartitioning(
                        neigh[0], neigh[1], neigh[2], neigh[3], neigh[4],
                        neigh[5], [])

                valid_partitioning = []
                for p in partitioning:
                    (y_min, y_max, x_min, x_max, z_min,
                     z_max) = (p[0], p[1], p[2], p[3], p[4], p[5])
                    failed_conditions = []
                    cond1 = utilityFunctions.hasValidGroundBlocks(
                        x_min, x_max, z_min, z_max, height_map)
                    if cond1 == False: failed_conditions.append(1)
                    cond2 = utilityFunctions.hasMinimumSize(
                        y_min, y_max, x_min, x_max, z_min, z_max, minimum_h,
                        minimum_w, mininum_d)
                    if cond2 == False: failed_conditions.append(2)
                    cond3 = utilityFunctions.hasAcceptableSteepness(
                        x_min, x_max, z_min, z_max, height_map,
                        utilityFunctions.getScoreArea_type1, threshold)
                    if cond3 == False: failed_conditions.append(3)
                    if cond1 and cond2 and cond3:
                        score = utilityFunctions.getScoreArea_type1(
                            height_map, x_min, x_max, z_min, z_max)
                        valid_partitioning.append((score, p))
                        logging.info("Passed the 3 conditions!")
                    else:
                        logging.info(
                            "Failed Conditions {}".format(failed_conditions))

                partitioning_list.extend(valid_partitioning)
                logging.info(
                    "Generated a partition with {} valid lots and {} invalids ones"
                    .format(len(valid_partitioning),
                            len(partitioning) - len(valid_partitioning)))

        temp_partitioning_list.extend(partitioning_list)
        # order partitions by steepness
        temp_partitioning_list = sorted(temp_partitioning_list)
        final_partitioning = utilityFunctions.getNonIntersectingPartitions(
            temp_partitioning_list)

        available_lots = len(final_partitioning)
        logging.info(
            "Current neighbourhood partitioning with most available_lots: {}, current threshold {}"
            .format(available_lots, threshold))

        threshold += 1
        current_try += 1

        logging.info("Final lots ({})for the neighbourhood {}: ".format(
            len(final_partitioning), neigh))
        for p in final_partitioning:
            logging.info("\t{}".format(p))

    for partition in final_partitioning:
        if well_count < 1 and biome in biome_with_well:
            well = generateWell(world, partition, height_map, biome)
            well_count += 1
            all_buildings.append(well)
        elif greenhouse_count * GREENHOUSE_CAPACITY < inhabitants:
            greenhouse = generateGreenhouse(world, partition, height_map,
                                            usable_wood, biome)
            greenhouse_count += 1
            all_buildings.append(greenhouse)
        else:
            house = generateHouse(world, partition, height_map, usable_wood,
                                  biome)
            inhabitants += RNG.randint(1, HOUSE_SIZE)
            all_buildings.append(house)

    # ==== GENERATE PATH MAP  ====
    # generate a path map that gives the cost of moving to each neighbouring cell
    pathMap = utilityFunctions.getPathMap(height_map, width, depth)

    # ==== CONNECTING BUILDINGS WITH ROADS  ====
    logging.info("Calling MST on {} buildings".format(len(all_buildings)))
    MST = utilityFunctions.getMST_Manhattan(all_buildings, pathMap, height_map)

    pavementBlockID = BlocksInfo.getPavmentId(biome)
    for m in MST:
        p1 = m[1]
        p2 = m[2]
        logging.info("Pavement with distance {} between {} and {}".format(
            m[0], p1.entranceLot, p2.entranceLot))

        path = utilityFunctions.aStar(p1.entranceLot, p2.entranceLot, pathMap,
                                      height_map)
        if path != None:
            logging.info(
                "Found path between {} and {}. Generating road...".format(
                    p1.entranceLot, p2.entranceLot))
            GeneratePath.generatPath(world, path, height_map, pavementBlockID)
        else:
            logging.info(
                "Couldnt find path between {} and {}. Generating a straight road between them..."
                .format(p1.entranceLot, p2.entranceLot))
            GeneratePath.generatPath_StraightLine(world, p1.entranceLot[1],
                                                  p1.entranceLot[2],
                                                  p2.entranceLot[1],
                                                  p2.entranceLot[2],
                                                  height_map, pavementBlockID)

    # ==== UPDATE WORLD ====
    world.updateWorld()