Perform a feature analysis

Learn how to perform a hot spot feature analysis with the spatial analysis service.

Run a find hot spot analysis on traffic and parking violations in San Francisco.

A feature analysis is the process of using the spatial analysis service to perform server-side geometric and analytic operations on feature data. All feature analysis requests are job requests. The easiest way to programmatically run an analysis request to the spatial analysis service is to use ArcGIS REST JS, which provides a Job class that handles long- running operations.

In this tutorial, you use ArcGIS REST JS to perform a hot spot analysis to find statistically significant clusters of parking violations.

Prerequisites

You need the following to access the spatial analysis service and perform feature analysis operations:

An ArcGIS Online account.
An OAuth2.0 credential to obtain its Client ID.
The redirect URL of the registered application to use for authentication in the format "https://<server>[:port]" as in: https://localhost:8080.

Steps

Get the starter code

Download the tutorial starter code zip file from the portal.
Unzip the folder and open it in a text editor of your choice, such as Visual Studio Code. The starter app includes the following:
- callback.html: This contains callback as part of the authentication process.
- index.html: This contains app logic and the OAuth 2.0 code necessary to perform the authentication.

Set up authentication

Create a new OAuth credential to register the application.

Go to the Create OAuth credentials for user authentication tutorial to get an OAuth credential.
Copy the Client ID and Redirect URL from your OAuth credentials item and paste them to a safe location. They will be used in a later step.

All users that access this application need an ArcGIS account with privileges to access the basemap styles service and spatial analysis service.

Set the developer credentials

In both the index.html and callback.html files, replace YOUR_CLIENT_ID and YOUR_REDIRECT_URL with the client ID and redirect URL of your OAuth credentials.

index.html
Use dark colors for code blocks
    // Your client ID from OAuth credentials
    const clientId = "YOUR_CLIENT_ID";
    // The redirect URL registered in your OAuth credentials
    const redirectUri = "YOUR_REDIRECT_URL";

callback.html
Use dark colors for code blocks
      // Your client ID from OAuth credentials
      const clientId = "YOUR_CLIENT_ID";
      // The redirect URL registered in your OAuth credentials
      const redirectUri = "YOUR_REDIRECT_URL";

Run the app and ensure you can sign in successfully.

If you are unable to sign in, make sure you have the correct redirect URL and port. This URL varies based on your application and typically takes the format of https://<server>[:port]/callback.html or http://my-arcgis-app:/auth. For example, if you are running an application on http://127.0.0.1:5500/, add http://127.0.0.1:5500/callback.html as your redirect URL.

Add script references

In addition to OpenLayers, also reference the Portal helper class from ArcGIS REST JS to access the spatial analysis service URL.

Add a reference to the ArcGIS REST JS Portal helper class.

Use dark colors for code blocks
  <!-- ArcGIS REST JS used for user authentication. -->
  <script src="https://unpkg.com/@esri/arcgis-rest-request@4.0.0/dist/bundled/request.umd.js"></script>
  <!-- OpenLayers -->
  <link rel="stylesheet" href="https://cdn.jsdelivr.net/npm/ol@v10.1.0/ol.css" type="text/css" />
  <script src="https://cdn.jsdelivr.net/npm/ol@v10.1.0/dist/ol.js"></script>
  <script src="https://cdn.jsdelivr.net/npm/ol-mapbox-style@12.3.5/dist/olms.js"></script>

  <!--ArcGIS REST JS used for spatial analysis -->
  <script src="https://unpkg.com/@esri/arcgis-rest-portal/dist/bundled/portal.umd.min.js"></script>

Update the map

Update the map's viewpoint to [37.760, -122.445] and a zoom level of 13 to focus in San Fransisco, CA. Then, update the basemap to arcgis/human-geography.

Expand
Use dark colors for code blocks
    // create map and add basemap
    const map = new ol.Map({ target: "map" });

    map.setView(
      new ol.View({

        center: ol.proj.fromLonLat([-122.445, 37.76]),
        zoom: 13

      })
    );
Expand

Display parking violations

To perform feature analysis, you need to provide feature data as input. In this tutorial, you use the SF parking violations hosted feature layer as input data for the hot spot analysis.

Add the parking violation feature layer and the corresponding vector tile layer to the points and pointsVTL variables.

Expand
Use dark colors for code blocks
      const points =
        "https://services3.arcgis.com/GVgbJbqm8hXASVYi/arcgis/rest/services/sf_traffic_parking_violations_sa_osapi/FeatureServer/0";

      const pointsVTL =
        "https://vectortileservices3.arcgis.com/GVgbJbqm8hXASVYi/arcgis/rest/services/sf_traffic_parking_violations_sa_osapi/VectorTileServer";
Expand

Add the SF parking violations vector tile layer to the map.

Expand
Use dark colors for code blocks

      const inputSource = new ol.source.VectorTile({
        format: new ol.format.MVT(),
        url: `${pointsVTL}/tile/{z}/{y}/{x}.pbf`
      });

      // add the input data to the map
      const vtlLayer = new ol.layer.VectorTile({
        source: inputSource,

        style: (f) => {
          return new ol.style.Style({
            image: new ol.style.Circle({
              fill: new ol.style.Fill({
                color: "rgb(0,0,0)"
              }),
              radius: 1.5
            })
          });
        }
      });

      map.addLayer(vtlLayer);

    });

Expand

Add the data attribution for the vector tile layer source.

Go to the sf_traffic_parking_violations_sa_osapi_vtl item.
Scroll down to the Credits (Attribution) section and copy its value.

Create an attributions property and paste the attribution value from the item.

Expand
Use dark colors for code blocks

      const inputSource = new ol.source.VectorTile({
        format: new ol.format.MVT(),
        url: `${pointsVTL}/tile/{z}/{y}/{x}.pbf`
      });

      // add the input data to the map
      const vtlLayer = new ol.layer.VectorTile({
        source: inputSource,

        // Attribution text retrieved from https://arcgis.com/home/item.html?id=a9b4d69b02ee4365a0fcb9610c69db9b
        attributions: ["| San Francisco 311, City and County of San Francisco"],

        style: (f) => {
          return new ol.style.Style({
            image: new ol.style.Circle({
              fill: new ol.style.Fill({
                color: "rgb(0,0,0)"
              }),
              radius: 1.5
            })
          });
        }
      });

      map.addLayer(vtlLayer);

    });

Expand

Run the application and navigate to your localhost, for example https://localhost:8080. After you sign in, you will see the vector tile layer rendered on the map.

Get the analysis URL

To make a request to the spatial analysis service, you need to get the URL first. The analysis service URL is unique to your organization.

Call the getSelf operation from ArcGIS REST JS to obtain the analysis URL.

Expand
Use dark colors for code blocks
    const getAnalysisUrl = async (withAuth) => {
      const portalSelf = await arcgisRest.getSelf({
        authentication: withAuth
      });

      return portalSelf.helperServices.analysis.url;
    };
Expand

Make the request

Use the Job class and set the operationURL and params required for the selected analysis.

Create a function that submits a Job request to the spatial analysis service using your organization's analysis URL. Set the params required for a hot spot analysis and authenticate using session.

Expand
Use dark colors for code blocks
    const runAnalysis = async () => {
      const analysisUrl = await getAnalysisUrl(session);
      const operationUrl = `${analysisUrl}/FindHotSpots/submitJob`;
      const params = {
        analysisLayer: {
          url: "https://services3.arcgis.com/GVgbJbqm8hXASVYi/arcgis/rest/services/sf_traffic_parking_violations_sa_osapi/FeatureServer/0"
        },
        shapeType: "Hexagon",
        outputName: {
          serviceProperties: {
            name: `OpenLayers_find_hot_spots_${new Date().getTime()}`
          }
        } //Outputs results as a hosted feature serivce.
      };

      const jobReq = await arcgisRest.Job.submitJob({
        url: operationUrl,
        params: params,
        authentication: session
      });

      // get all the results, this will start monitoring and trigger events
      const jobResp = await jobReq.getAllResults();

      // jobResp.aggregatedLayer.value.url
      return jobResp;
    };

Expand

Create an HTML div element called status and a loader to display the job status.

Use dark colors for code blocks

<body>

  <div id="status">
    <span class="loader"></span>
    <div id="info"></div>
  </div>

  <div id="map"></div>

Apply CSS styling to the status and the loader.

Expand
Use dark colors for code blocks
    #status {
      width: 400px;
      background-color: #303336;
      color: #edffff;
      z-index: 0;
      position: absolute;
      font-family: Arial, Helvetica, sans-serif;
      top: 50%;
      left: 50%;
      transform: translate(-50%, -50%);
      display: flex;
      flex-direction: row;
      justify-content: center;
      padding: 5px;
      border-radius: 5px;
      gap: 10px;
      align-content: center;
      align-items: center;
      margin: auto;
    }

    .loader {
      width: 24px;
      height: 24px;
      border: 5px solid #edffff;
      border-bottom-color: #303336;
      border-radius: 50%;
      display: inline-block;
      box-sizing: border-box;
      animation: rotation 2s linear infinite;
    }

    @keyframes rotation {
      0% {
        transform: rotate(0deg);
      }

      100% {
        transform: rotate(360deg);
      }
    }

Expand

Create a function called showInfo to control the visibility and content of the status element.

Expand
Use dark colors for code blocks
    const showInfo = (visible, msg) => {
      const statusUi = document.getElementById("status");
      !visible ? (statusUi.style.zIndex = 0) : (statusUi.style.zIndex = 1001);
      const infoUi = document.getElementById("info");
      infoUi.innerText = msg;
    };

    const runAnalysis = async () => {
      const analysisUrl = await getAnalysisUrl(session);
      const operationUrl = `${analysisUrl}/FindHotSpots/submitJob`;
      const params = {
        analysisLayer: {
          url: "https://services3.arcgis.com/GVgbJbqm8hXASVYi/arcgis/rest/services/sf_traffic_parking_violations_sa_osapi/FeatureServer/0"
        },
        shapeType: "Hexagon",
        outputName: {
          serviceProperties: {
            name: `OpenLayers_find_hot_spots_${new Date().getTime()}`
          }
        } //Outputs results as a hosted feature serivce.
      };

      const jobReq = await arcgisRest.Job.submitJob({
        url: operationUrl,
        params: params,
        authentication: session
      });

      // get all the results, this will start monitoring and trigger events
      const jobResp = await jobReq.getAllResults();

      // jobResp.aggregatedLayer.value.url
      return jobResp;
    };

Expand

Use the function to dynamically update the status element with the current status of the analysis.

Expand
Use dark colors for code blocks
    const runAnalysis = async () => {
      const analysisUrl = await getAnalysisUrl(session);
      const operationUrl = `${analysisUrl}/FindHotSpots/submitJob`;
      const params = {
        analysisLayer: {
          url: "https://services3.arcgis.com/GVgbJbqm8hXASVYi/arcgis/rest/services/sf_traffic_parking_violations_sa_osapi/FeatureServer/0"
        },
        shapeType: "Hexagon",
        outputName: {
          serviceProperties: {
            name: `OpenLayers_find_hot_spots_${new Date().getTime()}`
          }
        } //Outputs results as a hosted feature serivce.
      };

      const jobReq = await arcgisRest.Job.submitJob({
        url: operationUrl,
        params: params,
        authentication: session
      });

      showInfo(true, "Initializing analysis...");
      // listen to the status event to get updates every time the job status is checked.
      jobReq.on(arcgisRest.JOB_STATUSES.Status, (jobInfo) => {
        console.log(jobInfo.status);
        showInfo(true, "Hot spot analysis running...");
      });

      // get all the results, this will start monitoring and trigger events
      const jobResp = await jobReq.getAllResults();

      // jobResp.aggregatedLayer.value.url
      return jobResp;
    };
Expand

Call the runAnalysis function and store its response in jobResponse.

Expand
Use dark colors for code blocks
    const showInfo = (visible, msg) => {
      const statusUi = document.getElementById("status");
      !visible ? (statusUi.style.zIndex = 0) : (statusUi.style.zIndex = 1001);
      const infoUi = document.getElementById("info");
      infoUi.innerText = msg;
    };

    const runAnalysis = async () => {
      const analysisUrl = await getAnalysisUrl(session);
      const operationUrl = `${analysisUrl}/FindHotSpots/submitJob`;
      const params = {
        analysisLayer: {
          url: "https://services3.arcgis.com/GVgbJbqm8hXASVYi/arcgis/rest/services/sf_traffic_parking_violations_sa_osapi/FeatureServer/0"
        },
        shapeType: "Hexagon",
        outputName: {
          serviceProperties: {
            name: `OpenLayers_find_hot_spots_${new Date().getTime()}`
          }
        } //Outputs results as a hosted feature serivce.
      };

      const jobReq = await arcgisRest.Job.submitJob({
        url: operationUrl,
        params: params,
        authentication: session
      });

      showInfo(true, "Initializing analysis...");
      // listen to the status event to get updates every time the job status is checked.
      jobReq.on(arcgisRest.JOB_STATUSES.Status, (jobInfo) => {
        console.log(jobInfo.status);
        showInfo(true, "Hot spot analysis running...");
      });

      // get all the results, this will start monitoring and trigger events
      const jobResp = await jobReq.getAllResults();

      // jobResp.aggregatedLayer.value.url
      return jobResp;
    };

    const jobResults = await runAnalysis();
Expand

Display results

The results of a feature analysis are returned as feature data.

Create a new vector source object from the analysis results hosted feature layer URL as GeoJSON. As URL parameters, Add a query string to only include results with a value greater than 0, set the outFields, and provide the session token.

Expand
Use dark colors for code blocks

    showInfo(true, "Adding results to map...");
    const vectorSource = new ol.source.Vector({
      format: new ol.format.GeoJSON(),
      url: `${jobResults.hotSpotsResultLayer.value.url}/query?f=geojson&where=Gi_Bin<>0&outFields=Gi_Text&token=${session.token}`
    });

Expand

Create a new layer from the source, define it's style and add the layer to the map. Then, hide the status element.

Expand
Use dark colors for code blocks

    showInfo(true, "Adding results to map...");
    const vectorSource = new ol.source.Vector({
      format: new ol.format.GeoJSON(),
      url: `${jobResults.hotSpotsResultLayer.value.url}/query?f=geojson&where=Gi_Bin<>0&outFields=Gi_Text&token=${session.token}`
    });

    const resultLayer = new ol.layer.Vector({
      source: vectorSource,
      style: (feature) => {
        const fieldValue = feature.get("Gi_Text");
        const colorTable = {
          "Hot Spot with 99% Confidence": "#d62f27",
          "Hot Spot with 95% Confidence": "#ed7551",
          "Hot Spot with 90% Confidence": "#fab984",
          "Not Significant": "#f7f7f2",
          "Cold Spot with 90% Confidence": "#c0ccbe",
          "Cold Spot with 95% Confidence": "#849eba",
          "Cold Spot with 99% Confidence": "#4575b5"
        };

        return new ol.style.Style({
          fill: new ol.style.Fill({
            color: colorTable[fieldValue || "transparent"]
          })
        });
      }
    });
    map.addLayer(resultLayer);
    // hide the status indicator
    showInfo(false, "");
Expand

Run the app

Run the application and navigate to your localhost, for example https://localhost:8080.

After signing in, a request is sent to the spatial analysis service to perform a hot spot analysis. When the job is complete, the results of the analysis will display on the map. The analysis can take up to a minute to complete.