Find outliers

An outlier analysis resulting in a new layer with statistically significant clusters and anomalous population values by USA county.

What is an outlier analysis?

An outlier analysis is the process of identifying both clusters and anomalous values (outliers) in spatial data. It determines whether an attribute value or point count for each feature is significantly different, defined as the resultant z-score and p-value, from its neighbors. To execute the analysis, use the spatial analysis service and the FindOutliers operation.

The analysis classifies features as being:

High-High: a high value surrounded by other high values
High-Low: a high value surrounded by low values
Low-High: a low value surrounded by high values
Low-Low: a low value surrounded by other low values.

A feature is part of a cluster when it has a similar value to its neighbors. A feature is considered an outlier when it has dissimilar values from its neighbors.

An outlier analysis helps to find spatial trends and patterns in the data that may not be visible at first glance.

Real-world examples of this analysis include the following:

Finding outliers (either high or low counts) of traffic crashes or crimes.
Determining whether there are outlier (anomalous) spending trends or real estate prices.
Finding whether some areas of a country might have a higher population despite being surrounded by lower population numbers.

How to perform an outlier analysis

The general steps to performing an outlier analysis are as follows:

Review the parameters for the FindOutliers operation.
Send a request to get the spatial analysis service URL.
Execute a job request with the following URL and parameters:
- URL: https://<YOUR_ANALYSIS_SERVICE>/arcgis/rest/services/tasks/GPServer/FindOutliers/submitJob
- Parameters:
  - analysisLayer: Your dataset as a hosted feature layer or feature collection.
  - analysisField: A numeric field.
  - shapeType: Fishnet or Hexagon.
  - outputName: A string representing the name of the hosted feature layer to reutrn with the results.
Check the status.
Get the output layer results.

To see examples using ArcGIS API for Python, ArcGIS REST JS, and the ArcGIS REST API, go to Examples below.

URL request

Use dark colors for code blocksCopy
http://<YOUR_ANALYSIS_SERVICE>/arcgis/rest/services/tasks/GPServer/FindOutliers/submitJob?<parameters>

Required parameters

Name	Description	Examples
`f`	The format of the data returned.	`f=json` `f=pjson`
`token`	An OAuth 2.0 access token. Learn how to get an access token in Security and authentication.	`token=<ACCESS_TOKEN>`
`analysisLayer`	The point or polygon feature layer.	`{url:"https://services3.arcgis.com/GVgbJbqm8hXASVYi/arcgis/rest/services/Average_PDX_home_values/FeatureServer/0"}`
`analysisField`	Use if the analysis layer contains polygons.	`counts`, `rates`, `averages`

Key parameters

Name	Description	Examples
`shapeType`	Use if the analysis layer contains points.	`hexagon`, `fishnet`
`boundingPolygonLayer`	When the analysis layer contains points and no `analysisField` is specified, you can provide a boundary for the analysis.	`{url:"https://services3.arcgis.com/GVgbJbqm8hXASVYi/arcgis/rest/services/Delaware_County_Municipal_Boundaries/FeatureServer/0"}`
`outputName`	A string representing the name of the hosted feature layer to return with the results. NOTE: If you do not include this parameter, the results are returned as a feature collection (JSON).	`{"serviceProperties": {"name": "<SERVICE_NAME>"}}`
`context`	A bounding box or output spatial reference for the analysis.	`"extent":{"xmin:", "ymin:", "xmax:", "ymax:"}`

Code examples

Identify outliers in traffic crashes

This example uses the FindOutliers operation to determine where there are statistically significant outliers of Traffic crashes counted within a fishnet grid. The anomalous areas, shown as dark red and dark blue, indicate either significantly high or significantly low instances of crashes compared to neighboring clusters.

In the analysis, the analysisLayer value is the Traffic crashes hosted feature layer. The points in the layer are counted within a fishnet, which was set in the shapeType parameter.

Outlier analysis showing clusters and outliers of traffic crashes.

APIs

ArcGIS API for Python

ArcGIS REST JS

Expand
Use dark colors for code blocksCopy

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
trafficLayer = "https://services3.arcgis.com/GVgbJbqm8hXASVYi/arcgis/rest/services/Traffic_Crashes/FeatureServer/0"

results = find_outliers(
    analysis_layer=trafficLayer,
    shape_type="fishnet",
    #Outputs results as a hosted feature layer.
    output_name="Find outliers results"
)

result_features = results["outliers_result_layer"].layers[0].query()

print(
    f"The find outliers layer has {len(result_features.features)} new records"
)

Expand

Go to tutorial

Service requests

Request

HTTPHTTPcURL
Use dark colors for code blocksCopy
POST arcgis.com/sharing/rest/portals/self HTTP/1.1
Content-Type: application/x-www-form-urlencoded

&f=json
&token=<ACCESS_TOKEN>
Run in Postman

Response (JSON)

Use dark colors for code blocksCopy
{
  "helperServices": {
    // Other parameters
    "analysis": {
      "url": "https://<YOUR_ANALYSIS_SERVICE>/arcgis/rest/services/tasks/GPServer"
    },
    "geoenrichment": {
      "url": "https://geoenrich.arcgis.com/arcgis/rest/services/World/GeoenrichmentServer"
    }
  }
}

Request

HTTP

cURL

Use dark colors for code blocks

Copy

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
POST <YOUR_ANALYSIS_SERVICE>/arcgis/rest/services/tasks/GPServer/FindOutliers/submitJob HTTP/1.1
Content-Type: application/x-www-form-urlencoded

&f=json
&token=<ACCESS_TOKEN>
&analysisLayer={"url":"https://services3.arcgis.com/GVgbJbqm8hXASVYi/arcgis/rest/services/Traffic_Crashes/FeatureServer/0","name":"Traffic Crashes"}
&shapeType=fishnet
&returnProcessInfo=true
&outputName={"serviceProperties":{"name":"Find outliers results"}}
&context={}

Run in Postman

Response (JSON)

Use dark colors for code blocksCopy

{
  "jobId": "<JOB_ID>",
  "jobStatus": "esriJobSubmitted",
  "results": {},
  "inputs": {},
  "messages": []
}

Request

HTTP

cURL

Use dark colors for code blocks

Copy

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
POST <YOUR_ANALYSIS_SERVICE>/arcgis/rest/services/tasks/GPServer/FindOutliers/jobs/<JOB_ID> HTTP/1.1
Content-Type: application/x-www-form-urlencoded

&f=json
&token=<ACCESS_TOKEN>

Run in Postman

Response (JSON)

Use dark colors for code blocksCopy
// Executing job
{
  "jobId": "<JOB_ID>",
  "jobStatus": "esriJobExecuting",
  "results": {},
  "inputs": {},
  "messages": []
}
// Job succeeded
{
  "jobId": "<JOB_ID>",
  "jobStatus": "esriJobSucceeded",
  "results": {
      "outliersResultLayer": {
          "paramUrl": "results/outliersResultLayer"
      },
      "processInfo": {
          "paramUrl": "results/processInfo"
      }
  },
  "inputs": {},
  "messages": []
}

Request

HTTP

cURL

Use dark colors for code blocks

Copy

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
POST <YOUR_ANALYSIS_SERVICE>/arcgis/rest/services/tasks/GPServer/FindOutliers/jobs/<JOB_ID>/results/outliersResultLayer HTTP/1.1
Content-Type: application/x-www-form-urlencoded

&f=json
&returnType=data
&token=<ACCESS_TOKEN>

Run in Postman

Response (JSON)

Use dark colors for code blocksCopy
{
  "paramName": "outliersResultLayer",
  "dataType": "GPString",
  "value": {
      "url": "<SERVICE_URL>",
      "itemId": "<ITEM_ID>"
  }
}

Identify home value outliers

This example uses an outlier analysis to determine where there are statistically significant outliers for home values in Portland. The anomalous areas, shown as dark red and dark blue, indicate either significantly high or significantly low home values compared to neighboring clusters.

In the analysis, the analysisLayer value is the Enriched Portland hexagon bins hosted feature layer. The feature layer was created using generated hexagon bins that were enriched using data from the GeoEnrichment service. To analyze home values, you set the analysisField with the AVG_CY attribute.

To learn how to generate hexagon bins, go to generate tessellations.

Outlier analysis showing high and low home values.

APIs

ArcGIS API for Python

ArcGIS REST JS

Expand
Use dark colors for code blocksCopy

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
# https://developers.arcgis.com/rest/analysis/api-reference/feature-input.htm
pdx_homes= {
    "url": "https://services3.arcgis.com/GVgbJbqm8hXASVYi/arcgis/rest/services/Enriched_PDX_hex_bin_tessellations/FeatureServer/0",
    "filter": "AVGVAL_CY <> 0",
}

results = find_outliers(
    analysis_layer=pdx_homes,
    analysis_field="AVGVAL_CY",
    #Outputs results as a hosted feature layer.
    output_name="Find outliers results"
)

result_features = results["outliers_result_layer"].layers[0].query()

print(f"The outliers layer has {len(result_features.features)} new records")

Expand

Go to tutorial

Service requests

Request

HTTPHTTPcURL
Use dark colors for code blocksCopy
POST arcgis.com/sharing/rest/portals/self HTTP/1.1
Content-Type: application/x-www-form-urlencoded

&f=json
&token=<ACCESS_TOKEN>
Run in Postman

Response (JSON)

Use dark colors for code blocksCopy
{
  "helperServices": {
    // Other parameters
    "analysis": {
      "url": "https://<YOUR_ANALYSIS_SERVICE>/arcgis/rest/services/tasks/GPServer"
    },
    "geoenrichment": {
      "url": "https://geoenrich.arcgis.com/arcgis/rest/services/World/GeoenrichmentServer"
    }
  }
}

Request

HTTP

cURL

Use dark colors for code blocks

Copy

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35

POST <YOUR_ANALYSIS_SERVICE>/arcgis/rest/services/tasks/GPServer/FindOutliers/submitJob HTTP/1.1
Content-Type: application/x-www-form-urlencoded

&f=json
&token=<ACCESS_TOKEN>
&analysisLayer={"url":"https://services3.arcgis.com/GVgbJbqm8hXASVYi/arcgis/rest/services/Enriched_PDX_hex_bin_tessellations/FeatureServer/0","filter":"AVGVAL_CY <> 0"}
&analysisField=AVGVAL_CY
&returnProcessInfo=true
&outputName={"serviceProperties":{"name":"Find outliers results"}}
&context={"extent":{"xmin":-13701588.357587751,"ymin":5636873.660541155,"xmax":-13554599.952195555,"ymax":5730585.457218655,"spatialReference":{"wkid":102100,"latestWkid":3857}}}

Run in Postman

Response (JSON)

Use dark colors for code blocksCopy

{
  "jobId": "<JOB_ID>",
  "jobStatus": "esriJobSubmitted",
  "results": {},
  "inputs": {},
  "messages": []
}

Request

HTTP

cURL

Use dark colors for code blocks

Copy

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
POST <YOUR_ANALYSIS_SERVICE>/arcgis/rest/services/tasks/GPServer/FindOutliers/jobs/<JOB_ID> HTTP/1.1
Content-Type: application/x-www-form-urlencoded

&f=json
&token=<ACCESS_TOKEN>

Run in Postman

Response (JSON)

Use dark colors for code blocksCopy
// Executing job
{
  "jobId": "<JOB_ID>",
  "jobStatus": "esriJobExecuting",
  "results": {},
  "inputs": {},
  "messages": []
}
// Job succeeded
{
  "jobId": "<JOB_ID>",
  "jobStatus": "esriJobSucceeded",
  "results": {
      "outliersResultLayer": {
          "paramUrl": "results/outliersResultLayer"
      },
      "processInfo": {
          "paramUrl": "results/processInfo"
      }
  },
  "inputs": {},
  "messages": []
}

Request

HTTP

cURL

Use dark colors for code blocks

Copy

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
POST <YOUR_ANALYSIS_SERVICE>/arcgis/rest/services/tasks/GPServer/FindOutliers/jobs/<JOB_ID>/results/outliersResultLayer HTTP/1.1
Content-Type: application/x-www-form-urlencoded

&f=json
&returnType=data
&token=<ACCESS_TOKEN>

Run in Postman

Response (JSON)

Use dark colors for code blocksCopy
{
  "paramName": "outliersResultLayer",
  "dataType": "GPString",
  "value": {
      "url": "<SERVICE_URL>",
      "itemId": "<ITEM_ID>"
  }
}

Tutorials

Learn how to perform related analyses interactively with Map Viewer and programmatically with ArcGIS API for Python, ArcGIS REST JS, and ArcGIS REST API.

Discover patterns in data

Find patterns and trends in data using spatial analysis operations.

Services

Spatial analysis service

Process spatial datasets to discover relationships and patterns.

API support

	Find hot spots	Find outliers	Find point clusters	Calculate density	Interpolate points
ArcGIS Maps SDK for JavaScript	¹	¹	¹	¹	¹
ArcGIS Maps SDK for .NET
ArcGIS Maps SDK for Kotlin
ArcGIS Maps SDK for Swift
ArcGIS Maps SDK for Java
ArcGIS Maps SDK for Qt
ArcGIS API for Python
ArcGIS REST JS	¹	¹	¹	¹	¹
Esri Leaflet	²	²	²	²	²
MapBox GL JS	²	²	²	²	²
OpenLayers	²	²	²	²	²

Full supportPartial supportNo support

1. Access with geoprocessing task
2. Access via ArcGIS REST JS

Tools

Map Viewer

Create, explore, and share web maps for 2D applications.

ArcGIS Pro

Explore, visualize, and analyze both 2D and 3D data with desktop GIS tools.

Mapping APIs and location services

Find outliers

Was this page helpful?