Data bodu clusteringu v Android SDK

Při vizualizaci mnoha datových bodů na mapě se datové body mohou překrývat. Překrývání může způsobit, že se mapa stane nečitelným a obtížně použitelným. Data bodů clusteringu jsou proces kombinování dat bodů, která jsou blízko sebe, a jejich reprezentace na mapě jako jeden clusterovaný datový bod. Když se uživatel přibližuje k mapě, shluky se rozchýlí do jednotlivých datových bodů. Pokud pracujete s velkým počtem datových bodů, vylepšete uživatelské prostředí pomocí procesů clusteringu.


Požadavky

Nezapomeňte dokončit kroky v dokumentu Rychlý start: Vytvoření aplikace pro Android. Bloky kódu v tomto článku je možné vložit do obslužné rutiny onReady události maps.

Povolení clusteringu ve zdroji dat

Povolte clustering DataSource ve třídě nastavením možnosti na cluster true . Nastavte clusterRadius na výběr okolních bodů a zkombinuje je do clusteru. Hodnota je clusterRadius v pixelech. Slouží clusterMaxZoom k určení úrovně přiblížení, na které chcete zakázat logiku clusteringu. Tady je příklad povolení clusteringu ve zdroji dat.

//Create a data source and enable clustering.
DataSource source = new DataSource(
    //Tell the data source to cluster point data.
    cluster(true),

    //The radius in pixels to cluster points together.
    clusterRadius(45),

    //The maximum zoom level in which clustering occurs.
    //If you zoom in more than this, all points are rendered as symbols.
    clusterMaxZoom(15)
);
 //Create a data source and enable clustering.
val source = DataSource( 
    //Tell the data source to cluster point data.
    cluster(true),  

    //The radius in pixels to cluster points together.
    clusterRadius(45),  

    //The maximum zoom level in which clustering occurs.
    //If you zoom in more than this, all points are rendered as symbols.
    clusterMaxZoom(15)
)

Upozornění

Clustering funguje jenom s Point funkcemi. Pokud zdroj dat obsahuje vlastnosti jiných typů geometrie, jako je nebo LineString Polygon , dojde k chybě.

Tip

Pokud jsou dva datové body blízko sebe na zemi, je možné, že se cluster nikdy neodliší bez ohledu na to, jak blízko se uživatel přiblíží. Pokud chcete tento přístup řeší, můžete nastavit možnost zakázat clusterMaxZoom logiku clusteringu a jednoduše zobrazit vše.

Třída DataSource poskytuje také následující metody související s clusteringem.

Metoda Návratový typ Description
getClusterChildren(Feature clusterFeature) FeatureCollection Načte děti daného clusteru na další úrovni přiblížení. Těmito podřízenými třídami může být kombinace tvarů a podclusterů. Dílčí clustery budou funkce s vlastnostmi, které odpovídají vlastnosti ClusteredProperties.
getClusterExpansionZoom(Feature clusterFeature) int Vypočítá úroveň přiblížení, na které se cluster začne zvětšovat nebo rozdělit.
getClusterLeaves(Feature clusterFeature, long limit, long offset) FeatureCollection Načte všechny body v clusteru. Nastavte limit , aby se vracel podmnožina bodů, a pomocí použijte k offset prochádování bodů.

Zobrazení clusterů pomocí vrstvy bublin

Bublinová vrstva je skvělým způsobem, jak vykreslit shlukované body. Pomocí výrazů můžete škálovat poloměr a změnit barvu na základě počtu bodů v clusteru. Pokud clustery zobrazujete pomocí vrstvy bublin, měli byste k vykreslení neclusterovaných datových bodů použít samostatnou vrstvu.

Pokud chcete zobrazit velikost clusteru v horní části bubliny, použijte vrstvu symbolů s textem a nepoužívejte ikonu.

Následující kód zobrazí seskupené body pomocí vrstvy bublin a počet bodů v každém clusteru pomocí vrstvy symbolů. Druhá vrstva symbolů slouží k zobrazení jednotlivých bodů, které nejsou v rámci clusteru.

//Create a data source and add it to the map.
DataSource source = new DataSource(
    //Tell the data source to cluster point data.
    cluster(true),

    //The radius in pixels to cluster points together.
    clusterRadius(45),

    //The maximum zoom level in which clustering occurs.
    //If you zoom in more than this, all points are rendered as symbols.
    clusterMaxZoom(15)
);

//Import the geojson data and add it to the data source.
map.importDataFromUrl("https://earthquake.usgs.gov/earthquakes/feed/v1.0/summary/all_week.geojson");

//Add data source to the map.
map.sources.add(source);

//Create a bubble layer for rendering clustered data points.
map.layers.add(new BubbleLayer(source,
    //Scale the size of the clustered bubble based on the number of points in the cluster.
    bubbleRadius(
        step(
            get("point_count"),
            20,             //Default of 20 pixel radius.
            stop(100, 30),  //If point_count >= 100, radius is 30 pixels.
            stop(750, 40)   //If point_count >= 750, radius is 40 pixels.
        )
    ),

    //Change the color of the cluster based on the value on the point_cluster property of the cluster.
    bubbleColor(
        step(
            toNumber(get("point_count")),
            color(Color.GREEN),              //Default to lime green.
            stop(100, color(Color.YELLOW)),  //If the point_count >= 100, color is yellow.
            stop(750, color(Color.RED))      //If the point_count >= 100, color is red.
        )
    ),

    bubbleStrokeWidth(0f),

    //Only rendered data points which have a point_count property, which clusters do.
    BubbleLayerOptions.filter(has("point_count"))
));

//Create a symbol layer to render the count of locations in a cluster.
map.layers.add(new SymbolLayer(source,
    iconImage("none"),                //Hide the icon image.
    textField(get("point_count")),    //Display the point count as text.
    textOffset(new Float[]{ 0f, 0.4f }),

    //Allow clustered points in this layer.
    SymbolLayerOptions.filter(has("point_count"))
));

//Create a layer to render the individual locations.
map.layers.add(new SymbolLayer(source,
    //Filter out clustered points from this layer.
    SymbolLayerOptions.filter(not(has("point_count")))
));
//Create a data source and add it to the map.
val source = DataSource( 
    //Tell the data source to cluster point data.
    cluster(true),  

    //The radius in pixels to cluster points together.
    clusterRadius(45),  

    //The maximum zoom level in which clustering occurs.
    //If you zoom in more than this, all points are rendered as symbols.
    clusterMaxZoom(15)
)

//Import the geojson data and add it to the data source.
map.importDataFromUrl("https://earthquake.usgs.gov/earthquakes/feed/v1.0/summary/all_week.geojson")

//Add data source to the map.
map.sources.add(source)

//Create a bubble layer for rendering clustered data points.
map.layers.add(
    BubbleLayer(
        source,  

        //Scale the size of the clustered bubble based on the number of points in the cluster.
        bubbleRadius(
            step(
                get("point_count"),
                20,  //Default of 20 pixel radius.
                stop(100, 30),  //If point_count >= 100, radius is 30 pixels.
                stop(750, 40) //If point_count >= 750, radius is 40 pixels.
            )
        ),  

        //Change the color of the cluster based on the value on the point_cluster property of the cluster.
        bubbleColor(
            step(
                toNumber(get("point_count")),
                color(Color.GREEN),  //Default to lime green.
                stop(100, color(Color.YELLOW)),  //If the point_count >= 100, color is yellow.
                stop(750, color(Color.RED)) //If the point_count >= 100, color is red.
            )
        ),
        bubbleStrokeWidth(0f),  

        //Only rendered data points which have a point_count property, which clusters do.
        BubbleLayerOptions.filter(has("point_count"))
    )
)

//Create a symbol layer to render the count of locations in a cluster.
map.layers.add(
    SymbolLayer(
        source,
        iconImage("none"),  //Hide the icon image.
        textField(get("point_count")),  //Display the point count as text.
        textOffset(arrayOf(0f, 0.4f)),  

        //Allow clustered points in this layer.
        SymbolLayerOptions.filter(has("point_count"))
    )
)

//Create a layer to render the individual locations.
map.layers.add(
    SymbolLayer(
        source,  

        //Filter out clustered points from this layer.
        SymbolLayerOptions.filter(not(has("point_count")))
    )
)

Následující obrázek znázorňuje výše uvedený kód, který zobrazuje funkce seskupených bodů v bublinové vrstvě škálované a barevné na základě počtu bodů v clusteru. Neclusterované body se vykreslují pomocí vrstvy symbolů.

Rozdělení clusterovaných umístění map při přiblížení mapy

Zobrazení clusterů pomocí vrstvy symbolů

Při vizualizaci datových bodů vrstva symbolů automaticky skryje symboly, které se vzájemně překrývají, aby se zajistilo čistší uživatelské rozhraní. Toto výchozí chování může být nežádoucí, pokud chcete zobrazit hustotu datových bodů na mapě. Tato nastavení ale můžete změnit. Pokud chcete zobrazit všechny symboly, iconAllowOverlap nastavte možnost vrstvy Symbol na true .

Pomocí clusteringu můžete zobrazit hustotu datových bodů a zachovat čisté uživatelské rozhraní. Následující ukázka ukazuje, jak přidat vlastní symboly a reprezentovat shluky a jednotlivé datové body pomocí vrstvy symbolů.

//Load all the custom image icons into the map resources.
map.images.add("earthquake_icon", R.drawable.earthquake_icon);
map.images.add("warning_triangle_icon", R.drawable.warning_triangle_icon);

//Create a data source and add it to the map.
DataSource source = new DataSource(
    //Tell the data source to cluster point data.
    cluster(true)
);

//Import the geojson data and add it to the data source.
map.importDataFromUrl("https://earthquake.usgs.gov/earthquakes/feed/v1.0/summary/all_week.geojson");

//Add data source to the map.
map.sources.add(source);

//Create a symbol layer to render the clusters.
map.layers.add(new SymbolLayer(source,
    iconImage("warning_triangle_icon"),
    textField(get("point_count")),
    textOffset(new Float[]{ 0f, -0.4f }),

    //Allow clustered points in this layer.
    filter(has("point_count"))
));

//Create a layer to render the individual locations.
map.layers.add(new SymbolLayer(source,
    iconImage("earthquake_icon"),

    //Filter out clustered points from this layer.
    filter(not(has("point_count")))
));
//Load all the custom image icons into the map resources.
map.images.add("earthquake_icon", R.drawable.earthquake_icon)
map.images.add("warning_triangle_icon", R.drawable.warning_triangle_icon)

//Create a data source and add it to the map.
val source = DataSource( 
    //Tell the data source to cluster point data.
    cluster(true)
)

//Import the geojson data and add it to the data source.
map.importDataFromUrl("https://earthquake.usgs.gov/earthquakes/feed/v1.0/summary/all_week.geojson")

//Add data source to the map.
map.sources.add(source)

//Create a symbol layer to render the clusters.
map.layers.add(
    SymbolLayer(
        source,
        iconImage("warning_triangle_icon"),
        textField(get("point_count")),
        textOffset(arrayOf(0f, -0.4f)),  

        //Allow clustered points in this layer.
        filter(has("point_count"))
    )
)

//Create a layer to render the individual locations.
map.layers.add(
    SymbolLayer(
        source,
        iconImage("earthquake_icon"),  

        //Filter out clustered points from this layer.
        filter(not(has("point_count")))
    )
)

Pro tuto ukázku se do složky drawable aplikace načte následující obrázek.

Obrázek ikony zemětřesení Ikona počasí obrázek dešťových showerů
earthquake_icon.png warning_triangle_icon.png

Následující obrázek znázorňuje výše uvedený kód, který pomocí vlastních ikon vykresluje clusterované a nevyclusterované funkce bodů.

Mapa shlukovaných bodů vykreslených pomocí vrstvy symbolů

<a name="clustering-and-the-heat-maps-layer">Clustering a vrstva heat map

Heat mapy jsou skvělým způsobem, jak zobrazit hustotu dat na mapě. Tato metoda vizualizace sama o sobě dokáže zpracovat velký počet datových bodů. Pokud jsou datové body seskupené a velikost clusteru se používá jako váha heat mapy, může heat mapa zpracovat ještě více dat. Pokud chcete tuto možnost dosáhnout, heatmapWeight nastavte možnost vrstvy heat mapy na get(&quot;point_count") . Když je poloměr shluků malý, bude heat mapa vypadat téměř stejně jako heat mapa pomocí nevykluzovaných datových bodů, ale bude mít mnohem lepší výkon. Čím menší je ale poloměr clusteru, tím přesnější bude heat mapa, ale s menšími výhodami výkonu.

//Create a data source and add it to the map.
DataSource source = new DataSource(
    //Tell the data source to cluster point data.
    cluster(true),

    //The radius in pixels to cluster points together.
    clusterRadius(10)
);

//Import the geojson data and add it to the data source.
map.importDataFromUrl("https://earthquake.usgs.gov/earthquakes/feed/v1.0/summary/all_week.geojson");

//Add data source to the map.
map.sources.add(source);

//Create a heat map and add it to the map.
map.layers.add(new HeatMapLayer(source,
    //Set the weight to the point_count property of the data points.
    heatmapWeight(get("point_count")),

    //Optionally adjust the radius of each heat point.
    heatmapRadius(20f)
), "labels");
//Create a data source and add it to the map.
val source = DataSource( 
    //Tell the data source to cluster point data.
    cluster(true),  

    //The radius in pixels to cluster points together.
    clusterRadius(10)
)

//Import the geojson data and add it to the data source.
map.importDataFromUrl("https://earthquake.usgs.gov/earthquakes/feed/v1.0/summary/all_week.geojson")

//Add data source to the map.
map.sources.add(source)

//Create a heat map and add it to the map.
map.layers.add(
    HeatMapLayer(
        source,  

        //Set the weight to the point_count property of the data points.
        heatmapWeight(get("point_count")),  

        //Optionally adjust the radius of each heat point.
        heatmapRadius(20f)
    ), "labels"
)

Následující obrázek znázorňuje výše uvedený kód heat mapu, která je optimalizovaná pomocí funkcí seskupených bodů a počtu shluků jako váhy v heat mapě.

Mapa heat mapy optimalizované pomocí clusterovaných bodů jako váhy

Události myši v clusterovaných datových bodech

Když ve vrstvě, která obsahuje clusterované datové body, dojde k událostem myši, clusterovaný datový bod se k události vrátí jako objekt funkce bodu GeoJSON. Tato funkce bodu bude mít následující vlastnosti:

Název vlastnosti Typ Description
cluster boolean Určuje, jestli funkce představuje cluster.
point_count číslo Počet bodů, které cluster obsahuje.
point_count číslo Počet bodů, které cluster obsahuje.
point_count_abbreviated řetězec Řetězec, který zkřízuje point_count hodnotu , pokud je dlouhá. (například z 4 000 se stane 4K)

Tento příklad přebírá bublinovou vrstvu, která vykresluje body clusteru a přidává událost kliknutí. Když se událost kliknutí aktivuje, kód vypočítá a přiblíží mapu na další úroveň přiblížení, při které se cluster rozdělí. Tato funkce je implementována pomocí metody třídy a vlastnosti clusterového datového bodu, na který jste getClusterExpansionZoom DataSource cluster_id klikli.

//Create a data source and add it to the map.
DataSource source = new DataSource(
    //Tell the data source to cluster point data.
    cluster(true),

    //The radius in pixels to cluster points together.
    clusterRadius(45),

    //The maximum zoom level in which clustering occurs.
    //If you zoom in more than this, all points are rendered as symbols.
    clusterMaxZoom(15)
);

//Import the geojson data and add it to the data source.
source.importDataFromUrl("https://earthquake.usgs.gov/earthquakes/feed/v1.0/summary/all_week.geojson");

//Add data source to the map.
map.sources.add(source);

//Create a bubble layer for rendering clustered data points.
BubbleLayer clusterBubbleLayer = new BubbleLayer(source,
    //Scale the size of the clustered bubble based on the number of points in the cluster.
    bubbleRadius(
        step(
            get("point_count"),
            20f,    //Default of 20 pixel radius.
            stop(100, 30),   //If point_count >= 100, radius is 30 pixels.
            stop(750, 40)    //If point_count >= 750, radius is 40 pixels.
        )
    ),

    //Change the color of the cluster based on the value on the point_cluster property of the cluster.
    bubbleColor(
        step(
            get("point_count"),
            color(Color.GREEN),  //Default to green.
            stop(100, color(Color.YELLOW)), //If the point_count >= 100, color is yellow.
            stop(750, color(Color.RED))     //If the point_count >= 100, color is red.
        )
    ),

    bubbleStrokeWidth(0f),

    //Only rendered data points which have a point_count property, which clusters do.
    BubbleLayerOptions.filter(has("point_count"))   
);

//Add the clusterBubbleLayer and two additional layers to the map.
map.layers.add(clusterBubbleLayer);

//Create a symbol layer to render the count of locations in a cluster.
map.layers.add(new SymbolLayer(source,
    //Hide the icon image.
    iconImage("none"),   

    //Display the 'point_count_abbreviated' property value.
    textField(get("point_count_abbreviated")), 

    //Offset the text position so that it's centered nicely.
    textOffset(new Float[] { 0f, 0.4f }),

    //Only rendered data points which have a point_count property, which clusters do.
    SymbolLayerOptions.filter(has("point_count")) 
));

//Create a layer to render the individual locations.
map.layers.add(new SymbolLayer(source,
    //Filter out clustered points from this layer.
    SymbolLayerOptions.filter(not(has("point_count"))) 
));

//Add a click event to the cluster layer so we can zoom in when a user clicks a cluster.
map.events.add((OnFeatureClick) (features) -> {
    if(features.size() > 0) {
        //Get the clustered point from the event.
        Feature cluster = features.get(0);

        //Get the cluster expansion zoom level. This is the zoom level at which the cluster starts to break apart.
        int expansionZoom = source.getClusterExpansionZoom(cluster);

        //Update the map camera to be centered over the cluster.
        map.setCamera(
            //Center the map over the cluster points location.
            center((Point)cluster.geometry()),

            //Zoom to the clusters expansion zoom level.
            zoom(expansionZoom),

            //Animate the movement of the camera to the new position.
            animationType(AnimationType.EASE),
            animationDuration(200)
        );
    }

    //Return true indicating if event should be consumed and not passed further to other listeners registered afterwards, false otherwise.
    return true;
}, clusterBubbleLayer);
//Create a data source and add it to the map.
val source = DataSource( //Tell the data source to cluster point data.
    //The radius in pixels to cluster points together.
    cluster(true),  

    //The maximum zoom level in which clustering occurs.
    clusterRadius(45),  

    //If you zoom in more than this, all points are rendered as symbols.
    clusterMaxZoom(15)
)

//Import the geojson data and add it to the data source.
source.importDataFromUrl("https://earthquake.usgs.gov/earthquakes/feed/v1.0/summary/all_week.geojson")

//Add data source to the map.
map.sources.add(source)

//Create a bubble layer for rendering clustered data points.
val clusterBubbleLayer = BubbleLayer(
    source,  

    //Scale the size of the clustered bubble based on the number of points in the cluster.
    bubbleRadius(
        step(
            get("point_count"),
            20f,  //Default of 20 pixel radius.
            stop(100, 30),  //If point_count >= 100, radius is 30 pixels.
            stop(750, 40) //If point_count >= 750, radius is 40 pixels.
        )
    ),  

    //Change the color of the cluster based on the value on the point_cluster property of the cluster.
    bubbleColor(
        step(
            get("point_count"),
            color(Color.GREEN),  //Default to green.
            stop(
                100,
                color(Color.YELLOW)
            ),  //If the point_count >= 100, color is yellow.
            stop(750, color(Color.RED)) //If the point_count >= 100, color is red.
        )
    ),

    bubbleStrokeWidth(0f),  

    //Only rendered data points which have a point_count property, which clusters do.
    BubbleLayerOptions.filter(has("point_count"))
)

//Add the clusterBubbleLayer and two additional layers to the map.
map.layers.add(clusterBubbleLayer)

//Create a symbol layer to render the count of locations in a cluster.
map.layers.add(
    SymbolLayer(
        source,  

        //Hide the icon image.
        iconImage("none"),  

        //Display the 'point_count_abbreviated' property value.
        textField(get("point_count_abbreviated")),  

        //Offset the text position so that it's centered nicely.
        textOffset(
            arrayOf(
                0f,
                0.4f
            )
        ),  

        //Only rendered data points which have a point_count property, which clusters do.
        SymbolLayerOptions.filter(has("point_count"))
    )
)

//Create a layer to render the individual locations.
map.layers.add(
    SymbolLayer(
        source,  

        //Filter out clustered points from this layer.
        SymbolLayerOptions.filter(not(has("point_count")))
    )
)

//Add a click event to the cluster layer so we can zoom in when a user clicks a cluster.
map.events.add(OnFeatureClick { features: List<Feature?>? -> 
    if (features.size() > 0) {
        //Get the clustered point from the event.
        val cluster: Feature = features.get(0)

        //Get the cluster expansion zoom level. This is the zoom level at which the cluster starts to break apart.
        val expansionZoom: Int = source.getClusterExpansionZoom(cluster)

        //Update the map camera to be centered over the cluster.
        map.setCamera( 

            //Center the map over the cluster points location.
            center(cluster.geometry() as Point?),  

            //Zoom to the clusters expansion zoom level.
            zoom(expansionZoom),  

            //Animate the movement of the camera to the new position.
            animationType(AnimationType.EASE),
            animationDuration(200)
        )
    }
    true
}, clusterBubbleLayer)

Následující obrázek znázorňuje výše uvedený kód, který na mapě zobrazuje shlukované body, které při kliknutí přibližují další úroveň přiblížení, kterou se cluster začne rozebrat a rozbaluje.

Mapa clusterovaných funkcí, které se při kliknutí přibližují a rozlišují

Zobrazení oblasti clusteru

Data bodů, která cluster představuje, jsou rozložena do oblasti. V této ukázce dojde při najetí myší na cluster ke dvěma hlavním chováním. Nejprve se jednotlivé datové body obsažené v clusteru použije k výpočtu konvencí. Pak se na mapě zobrazí konventrní teplota, která zobrazí oblast. Konvexní smyšlový tvar je mnohoúhelník, který zabalí sadu bodů, jako je elastický pás, a lze ji vypočítat pomocí atlas.math.getConvexHull metody . Všechny body obsažené v clusteru lze načíst ze zdroje dat pomocí getClusterLeaves metody .

//Create a data source and add it to the map.
DataSource source = new DataSource(
    //Tell the data source to cluster point data.
    cluster(true)
);

//Import the geojson data and add it to the data source.
source.importDataFromUrl("https://earthquake.usgs.gov/earthquakes/feed/v1.0/summary/all_week.geojson");

//Add data source to the map.
map.sources.add(source);

//Create a data source for the convex hull polygon. Since this will be updated frequently it is more efficient to separate this into its own data source.
DataSource polygonDataSource = new DataSource();

//Add data source to the map.
map.sources.add(polygonDataSource);

//Add a polygon layer and a line layer to display the convex hull.
map.layers.add(new PolygonLayer(polygonDataSource));
map.layers.add(new LineLayer(polygonDataSource));

//Create a symbol layer to render the clusters.
SymbolLayer clusterLayer = new SymbolLayer(source,
    iconImage("marker-red"),
    textField(get("point_count_abbreviated")),
    textOffset(new Float[] { 0f, -1.2f }),
    textColor(Color.WHITE),
    textSize(14f),

    //Only rendered data points which have a point_count property, which clusters do.
    SymbolLayerOptions.filter(has("point_count"))
);
map.layers.add(clusterLayer);

//Create a layer to render the individual locations.
map.layers.add(new SymbolLayer(source,
    //Filter out clustered points from this layer.
    SymbolLayerOptions.filter(not(has("point_count")))
));

//Add a click event to the layer so we can calculate the convex hull of all the points within a cluster.
map.events.add((OnFeatureClick) (features) -> {
    if(features.size() > 0) {
        //Get the clustered point from the event.
        Feature cluster = features.get(0);

        //Get all points in the cluster. Set the offset to 0 and the max long value to return all points.
        FeatureCollection leaves = source.getClusterLeaves(cluster, Long.MAX_VALUE, 0);

        //Get the point features from the feature collection.
        List<Feature> childFeatures = leaves.features();

        //When only two points in a cluster. Render a line.
        if(childFeatures.size() == 2){
            //Extract the geometry points from the child features.
            List<Point> points = new ArrayList();

            childFeatures.forEach(f -> {
                points.add((Point)f.geometry());
            });

            //Create a line from the points.
            polygonDataSource.setShapes(LineString.fromLngLats(points));
        } else {
            Polygon hullPolygon = MapMath.getConvexHull(leaves);

            //Overwrite all data in the polygon data source with the newly calculated convex hull polygon.
            polygonDataSource.setShapes(hullPolygon);
        }
    }

    //Return true indicating if event should be consumed and not passed further to other listeners registered afterwards, false otherwise.
    return true;
}, clusterLayer);
//Create a data source and add it to the map.
val source = DataSource( 
    //Tell the data source to cluster point data.
    cluster(true)
)

//Import the geojson data and add it to the data source.
source.importDataFromUrl("https://earthquake.usgs.gov/earthquakes/feed/v1.0/summary/all_week.geojson")

//Add data source to the map.
map.sources.add(source)

//Create a data source for the convex hull polygon. Since this will be updated frequently it is more efficient to separate this into its own data source.
val polygonDataSource = DataSource()

//Add data source to the map.
map.sources.add(polygonDataSource)

//Add a polygon layer and a line layer to display the convex hull.
map.layers.add(PolygonLayer(polygonDataSource))
map.layers.add(LineLayer(polygonDataSource))

//Create a symbol layer to render the clusters.
val clusterLayer = SymbolLayer(
    source,
    iconImage("marker-red"),
    textField(get("point_count_abbreviated")),
    textOffset(arrayOf(0f, -1.2f)),
    textColor(Color.WHITE),
    textSize(14f),  

    //Only rendered data points which have a point_count property, which clusters do.
    SymbolLayerOptions.filter(has("point_count"))
)
map.layers.add(clusterLayer)

//Create a layer to render the individual locations.
map.layers.add(
    SymbolLayer(
        source,  

        //Filter out clustered points from this layer.
        SymbolLayerOptions.filter(not(has("point_count")))
    )
)

//Add a click event to the layer so we can calculate the convex hull of all the points within a cluster.
map.events.add(OnFeatureClick { features: List<Feature?>? -> 
    if (features.size() > 0) {
        //Get the clustered point from the event.
        val cluster: Feature = features.get(0)

        //Get all points in the cluster. Set the offset to 0 and the max long value to return all points.
        val leaves: FeatureCollection = source.getClusterLeaves(cluster, Long.MAX_VALUE, 0)

        //Get the point features from the feature collection.
        val childFeatures = leaves.features()

        //When only two points in a cluster. Render a line.
        if (childFeatures!!.size == 2) {
            //Extract the geometry points from the child features.
            val points: MutableList<Point?> = ArrayList()
            childFeatures!!.forEach(Consumer { f: Feature ->
                points.add(
                    f.geometry() as Point?
                )
            })

            //Create a line from the points.
            polygonDataSource.setShapes(LineString.fromLngLats(points))
        } else {
            val hullPolygon: Polygon = MapMath.getConvexHull(leaves)

            //Overwrite all data in the polygon data source with the newly calculated convex hull polygon.
            polygonDataSource.setShapes(hullPolygon)
        }
    }
    true
}, clusterLayer)

Následující obrázek znázorňuje výše uvedený kód, který zobrazuje oblast všech bodů v rámci shlukového kliknutí.

Mapa znázorňující konvexní mnohoúhelník všech bodů v rámci klikalého clusteru

Agregace dat v clusterech

Shluky se často reprezentují pomocí symbolu s počtem bodů, které jsou v rámci clusteru. Někdy je ale žádoucí přizpůsobit styl clusterů pomocí dalších metrik. S vlastnostmi clusteru je možné vytvořit vlastní vlastnosti, které se rovnají výpočtu na základě vlastností v rámci každého bodu s clusterem. Vlastnosti clusteru je možné definovat clusterProperties pomocí možnosti DataSource .

Následující kód vypočítá počet na základě vlastnosti typu entity každého datového bodu v clusteru. Když uživatel klikne na cluster, zobrazí se automaticky otevírané okno s dalšími informacemi o clusteru.

//An array of all entity type property names in features of the data set.
String[] entityTypes = new String[] { "Gas Station", "Grocery Store", "Restaurant", "School" };

//Create a popup and add it to the map.
Popup popup = new Popup();
map.popups.add(popup);

//Close the popup initially.
popup.close();

//Create a data source and add it to the map.
source = new DataSource(
    //Tell the data source to cluster point data.
    cluster(true),

    //The radius in pixels to cluster points together.
    clusterRadius(50),

    //Calculate counts for each entity type in a cluster as custom aggregate properties.
    clusterProperties(new ClusterProperty[]{
        new ClusterProperty("Gas Station", sum(accumulated(), get("Gas Station")), switchCase(eq(get("EntityType"), literal("Gas Station")), literal(1), literal(0))),
        new ClusterProperty("Grocery Store", sum(accumulated(), get("Grocery Store")), switchCase(eq(get("EntityType"), literal("Grocery Store")), literal(1), literal(0))),
        new ClusterProperty("Restaurant", sum(accumulated(), get("Restaurant")), switchCase(eq(get("EntityType"), literal("Restaurant")), literal(1), literal(0))),
        new ClusterProperty("School", sum(accumulated(), get("School")), switchCase(eq(get("EntityType"), literal("School")), literal(1), literal(0)))
    })
);

//Import the geojson data and add it to the data source.
source.importDataFromUrl("https://azuremapscodesamples.azurewebsites.net/Common/data/geojson/SamplePoiDataSet.json");

//Add data source to the map.
map.sources.add(source);

//Create a bubble layer for rendering clustered data points.
BubbleLayer clusterBubbleLayer = new BubbleLayer(source,
    bubbleRadius(20f),
    bubbleColor("purple"),
    bubbleStrokeWidth(0f),

    //Only rendered data points which have a point_count property, which clusters do.
    BubbleLayerOptions.filter(has("point_count"))
);

//Add the clusterBubbleLayer and two additional layers to the map.
map.layers.add(clusterBubbleLayer);

//Create a symbol layer to render the count of locations in a cluster.
map.layers.add(new SymbolLayer(source,
    //Hide the icon image.
    iconImage("none"),

    //Display the 'point_count_abbreviated' property value.
    textField(get("point_count_abbreviated")),

    textColor(Color.WHITE),
    textOffset(new Float[] { 0f, 0.4f }),

    //Only rendered data points which have a point_count property, which clusters do.
    SymbolLayerOptions.filter(has("point_count"))
));

//Create a layer to render the individual locations.
map.layers.add(new SymbolLayer(source,
    //Filter out clustered points from this layer.
    SymbolLayerOptions.filter(not(has("point_count")))
));

//Add a click event to the cluster layer and display the aggregate details of the cluster.
map.events.add((OnFeatureClick) (features) -> {
    if(features.size() > 0) {
        //Get the clustered point from the event.
        Feature cluster = features.get(0);

        //Create a number formatter that removes decimal places.
        NumberFormat nf = DecimalFormat.getInstance();
        nf.setMaximumFractionDigits(0);

        //Create the popup's content.
        StringBuilder sb = new StringBuilder();

        sb.append("Cluster size: ");
        sb.append(nf.format(cluster.getNumberProperty("point_count")));
        sb.append(" entities\n");

        for(int i = 0; i < entityTypes.length; i++) {
            sb.append("\n");

            //Get the entity type name.
            sb.append(entityTypes[i]);
            sb.append(": ");

            //Get the aggregated entity type count from the properties of the cluster by name.
            sb.append(nf.format(cluster.getNumberProperty(entityTypes[i])));
        }

        //Retrieve the custom layout for the popup.
        View customView = LayoutInflater.from(this).inflate(R.layout.popup_text, null);

        //Access the text view within the custom view and set the text to the title property of the feature.
        TextView tv = customView.findViewById(R.id.message);
        tv.setText(sb.toString());

        //Get the position of the cluster.
        Position pos = MapMath.getPosition((Point)cluster.geometry());

        //Set the options on the popup.
        popup.setOptions(
            //Set the popups position.
            position(pos),

            //Set the anchor point of the popup content.
            anchor(AnchorType.BOTTOM),

            //Set the content of the popup.
            content(customView)
        );

        //Open the popup.
        popup.open();
    }

    //Return a boolean indicating if event should be consumed or continue bubble up.
    return true;
}, clusterBubbleLayer);
//An array of all entity type property names in features of the data set.
val entityTypes = arrayOf("Gas Station", "Grocery Store", "Restaurant", "School")

//Create a popup and add it to the map.
val popup = Popup()
map.popups.add(popup)

//Close the popup initially.
popup.close()

//Create a data source and add it to the map.
val source = DataSource( 
    //Tell the data source to cluster point data.
    cluster(true),  

    //The radius in pixels to cluster points together.
    clusterRadius(50),  

    //Calculate counts for each entity type in a cluster as custom aggregate properties.
    clusterProperties(
        arrayOf<ClusterProperty>(
            ClusterProperty("Gas Station", sum(accumulated(), get("Gas Station")), switchCase(eq(get("EntityType"), literal("Gas Station")), literal(1), literal(0))),
            ClusterProperty("Grocery Store", sum(accumulated(), get("Grocery Store")), switchCase(eq(get("EntityType"), literal("Grocery Store")), literal(1), literal(0))),
            ClusterProperty("Restaurant", sum(accumulated(), get("Restaurant")), switchCase(eq(get("EntityType"), literal("Restaurant")), literal(1), literal(0))),
            ClusterProperty("School", sum(accumulated(), get("School")), switchCase(eq(get("EntityType"), literal("School")), literal(1), literal(0)))
        )
    )
)

//Import the geojson data and add it to the data source.
source.importDataFromUrl("https://azuremapscodesamples.azurewebsites.net/Common/data/geojson/SamplePoiDataSet.json")

//Add data source to the map.
map.sources.add(source)

//Create a bubble layer for rendering clustered data points.
val clusterBubbleLayer = BubbleLayer(
    source,
    bubbleRadius(20f),
    bubbleColor("purple"),
    bubbleStrokeWidth(0f),  

    //Only rendered data points which have a point_count property, which clusters do.
    BubbleLayerOptions.filter(has("point_count"))
)

//Add the clusterBubbleLayer and two additional layers to the map.
map.layers.add(clusterBubbleLayer)

//Create a symbol layer to render the count of locations in a cluster.
map.layers.add(
    SymbolLayer(
        source,  

        //Hide the icon image.
        iconImage("none"),  

        //Display the 'point_count_abbreviated' property value.
        textField(get("point_count_abbreviated")),

        textColor(Color.WHITE),
        textOffset(arrayOf(0f, 0.4f)),  

        //Only rendered data points which have a point_count property, which clusters do.
        SymbolLayerOptions.filter(has("point_count"))
    )
)

//Create a layer to render the individual locations.
map.layers.add(
    SymbolLayer(
        source,  

        //Filter out clustered points from this layer.
        SymbolLayerOptions.filter(not(has("point_count")))
    )
)

//Add a click event to the cluster layer and display the aggregate details of the cluster.
map.events.add(OnFeatureClick { features: List<Feature> ->
    if (features.size > 0) {
        //Get the clustered point from the event.
        val cluster = features[0]

        //Create a number formatter that removes decimal places.
        val nf: NumberFormat = DecimalFormat.getInstance()
        nf.setMaximumFractionDigits(0)

        //Create the popup's content.
        val sb = StringBuilder()

        sb.append("Cluster size: ")
        sb.append(nf.format(cluster.getNumberProperty("point_count")))
        sb.append(" entities\n")

        for (i in entityTypes.indices) {
            sb.append("\n")

            //Get the entity type name.
            sb.append(entityTypes[i])
            sb.append(": ")

            //Get the aggregated entity type count from the properties of the cluster by name.
            sb.append(nf.format(cluster.getNumberProperty(entityTypes[i])))
        }

        //Retrieve the custom layout for the popup.
        val customView: View = LayoutInflater.from(this).inflate(R.layout.popup_text, null)

        //Access the text view within the custom view and set the text to the title property of the feature.
        val tv: TextView = customView.findViewById(R.id.message)
        tv.text = sb.toString()

        //Get the position of the cluster.
        val pos: Position = MapMath.getPosition(cluster.geometry() as Point?)

        //Set the options on the popup.
        popup.setOptions( 
            //Set the popups position.
            position(pos),  

            //Set the anchor point of the popup content.
            anchor(AnchorType.BOTTOM),  

            //Set the content of the popup.
            content(customView)
        )

        //Open the popup.
        popup.open()
    }

    //Return a boolean indicating if event should be consumed or continue bubble up.
    true
} as OnFeatureClick, clusterBubbleLayer)

Automaticky otevíraná okna se řídí postupem popsaným v místním dokumentu zobrazení .

Následující obrázek ukazuje, že výše uvedený kód zobrazuje automaticky otevírané okno s agregovanými počty jednotlivých typů hodnot entit pro všechny body v kliknutí na clusterovaný bod.

Mapa znázorňující automaticky otevíraná okna agregovaných počtů typů entit všech bodů v clusteru

Další kroky

Přidání dalších dat do mapy: