Part 1 - Introduction to GeoEnrichment¶

1 Introduction
2 Getting Started
- 2.1 Object Model Diagram
- 2.2 Ways to enrich your data
3 Quick Example
4 Conclusion

Introduction¶

GeoEnrichment adds location intelligence to the data by providing facts about a location or an area. Using GeoEnrichment, you can get information about the people and places in a specific area or within a certain distance or drive time from a location. It enables you to query and use information from a large collection of data sets including population, income, housing, consumer behavior, and the natural environment. GeoEnrichment enables you to answer questions about locations that you can't answer with maps alone. For example: What kind of people live here? What do people like to do in this area? What are their habits and lifestyles?

GeoEnrichment makes your analysis more powerful by adding global demographic, spending, lifestyle or business features at different geographical levels such as city, county, region, state and country. Demographic features (Population, Age, Education etc.) and Socio-economic features (Income, Education, Wealth etc.) can be easily added to your location data, making it more intelligent. Feature Engineering is one of the key aspects of any Data Science project as it involves adding new features to the data to increase the predictive power of a learning algorithm. With GeoEnrichment, you can quickly add more features to your location data, helping your algorithms make better predictions.

To understand how GeoEnrichment adds value, let's imagine that a retail giant is evaluating potential sites to open new stores where the conditions for evaluation include competition, traffic, economic feasibility and market potential of different geographic areas. With GeoEnrichment, they can dig into at an average shoppers' lifestyle, income, spending, education and other socio-demographic factors for different neighborhoods to understand their potential customers and make an educated decision when choosing new sites.

Getting Started¶

A user must be logged on to a GIS in order to use GeoEnrichment. Geoenrichment functionality is available in the arcgis.geoenrichment module.

To enable GeoEnrichment, an ArcGIS Online subscription is needed or ArcGIS Enterprise needs to be configured with GeoEnrichment utility service. GeoEnrichment operations consume credits. Credits are the currency used across ArcGIS and are consumed for specific transactions. Learn more about credit consumption for GeoEnrichment here.

Object Model Diagram¶

The picture below illustrates how geoenrichment module is organized.

Ways to enrich your data¶

You can enrich your data in 2 ways:

enrich() method from arcgis.geoenrichment module.
enrich_layer( ) method from the features module.

The enrich() method returns a Spatiallly Enabled Data Frame. This data frame can be saved as a new feature layer Item in your GIS and used for analysis or visualization on a map. However, if you would like to enrich an existing FeatureLayer, then use the enrich_layer() method from the arcgis.features module. The result will be a new layer of input features that includes enriched data.

Quick Example¶

Let's look at a simple example of GeoEnrichment in action. Suppose a company wants to open a healthcare facility somewhere in Los Angeles, CA. They have a sample dataset of existing healthcare providers with their address details for the target areas (represented by their zip codes). The company wants to understand the demographics of each zip code to make the right decision.

Let's import this data and make it richer with GeoEnrichment.

In [1]:

# Import Libraries
import pandas as pd
from arcgis.gis import GIS
from arcgis.geoenrichment import Country

In [3]:

# Create a GIS Connection
gis = GIS(profile='your_online_profile')

In [5]:

# Read the data
df = pd.read_csv('../data/health.csv')
df

Out[5]:

	Number of Beds	Name	Address	City	State	Zip Code
0	156	Facility 1	2468 SOUTH ST ANDREWS PLACE	LOS ANGELES	CA	90018
1	59	Facility 2	2300 W. WASHINGTON BLVD.	LOS ANGELES	CA	90018
2	25	Facility 3	4060 E. WHITTIER BLVD.	LOS ANGELES	CA	90023
3	49	Facility 4	6070 W. PICO BOULEVARD	LOS ANGELES	CA	90035
4	55	Facility 5	1480 S. LA CIENEGA BL	LOS ANGELES	CA	90035

This dataset shows 5 providers with their address details. The providers are located in Zip Codes 90018, 90023 and 90035.

Let's enrich this dataset with socio-demographic factors such as Total Population, Median Age, Median Household Income, Diversity Index, Education for each zip code to better understand these areas.

In [6]:

# Define Analysis variables
analysis_variables = [
    'TOTPOP_CY',  # Population: Total Population (Esri)
    'DIVINDX_CY', # Diversity Index (Esri)
    'AVGHHSZ_CY', # Average Household Size (Esri)
    'MEDAGE_CY',  # Age: Median Age (Esri)
    'MEDHINC_CY', # Income: Median Household Income (Esri)
    'BACHDEG_CY', # Education: Bachelor's Degree (Esri)
]

In [7]:

# Get enriched data for each zip code
from arcgis.geoenrichment import *

usa = Country.get('US')
zip1 = usa.subgeographies.states['California'].zip5['90018']
zip2 = usa.subgeographies.states['California'].zip5['90023']
zip3 = usa.subgeographies.states['California'].zip5['90035']

enrich_df = enrich(study_areas=[zip1, zip2, zip3], analysis_variables=analysis_variables)

enrich_df

Out[7]:

	ID	OBJECTID	StdGeographyLevel	StdGeographyName	StdGeographyID	sourceCountry	aggregationMethod	populationToPolygonSizeRating	apportionmentConfidence	HasData	TOTPOP_CY	DIVINDX_CY	AVGHHSZ_CY	MEDAGE_CY	MEDHINC_CY	BACHDEG_CY	SHAPE
0	0	1	US.ZIP5	Los Angeles	90018	US	Query:US.ZIP5	2.191	2.576	1	52420	91.6	3.16	34.0	42741	4996	{"rings": [[[-118.30899000030098, 34.039920000...
1	1	2	US.ZIP5	Los Angeles	90023	US	Query:US.ZIP5	2.191	2.576	1	48673	76.9	4.29	29.1	43056	1367	{"rings": [[[-118.2062542231383, 34.0348220536...
2	2	3	US.ZIP5	Los Angeles	90035	US	Query:US.ZIP5	2.191	2.576	1	30187	59.0	2.25	38.8	88405	7848	{"rings": [[[-118.37619999961632, 34.059440000...

In [9]:

# Merge provider data with GeoEnrichment data
df['Zip Code'] = df['Zip Code'].apply(str)
merged = pd.merge(enrich_df, df, left_on='StdGeographyID',right_on='Zip Code')

In [13]:

merged.iloc[:,-10:]

Out[13]:

	MEDAGE_CY	MEDHINC_CY	BACHDEG_CY	SHAPE	Number of Beds	Name	Address	City	State	Zip Code
0	34.0	42741	4996	{'rings': [[[-118.30899000030098, 34.039920000...	156	Facility 1	2468 SOUTH ST ANDREWS PLACE	LOS ANGELES	CA	90018
1	34.0	42741	4996	{'rings': [[[-118.30899000030098, 34.039920000...	59	Facility 2	2300 W. WASHINGTON BLVD.	LOS ANGELES	CA	90018
2	29.1	43056	1367	{'rings': [[[-118.2062542231383, 34.0348220536...	25	Facility 3	4060 E. WHITTIER BLVD.	LOS ANGELES	CA	90023
3	38.8	88405	7848	{'rings': [[[-118.37619999961632, 34.059440000...	49	Facility 4	6070 W. PICO BOULEVARD	LOS ANGELES	CA	90035
4	38.8	88405	7848	{'rings': [[[-118.37619999961632, 34.059440000...	55	Facility 5	1480 S. LA CIENEGA BL	LOS ANGELES	CA	90035

Visualize on a map

Let's visualize the 3 zip codes on a map.

In [11]:

map1 = gis.map('Los Angeles, CA',12)
map1

ArcGIS API for Python Guide