ALL OVER THE MAP: COMPARING THE ACCURACY OF …...Accurate geocoding of a small record batch. Free...

The picture can't be displayed.

Robert Saul, EMI Consulting

Brian Billing, AEP Ohio

August 21, 2019

1

A L L O V E R T H E M A P : C O M P A R I N G T H E A C C U R A C Y O F G E O C O D I N G S O U R C E S

2 0 1 9 I E P E C C O N F E R E N C E P R E S E N T A T I O N

A G E N D A


1. “Why Does it Matter?”

2. Introduction

3. Methods

4. Results

5. Conclusion

“ W H Y D O E S I T M A T T E R ? ”B E N E F I T S O F S P A T I A L A N A L Y S I S I N E E P R O G R A M S

M I C R O - T A R G E T I N G S P A T I A L P A T T E R N S S E G M E N T A T I O N


As energy efficiency programs become more successful, utilities are more frequently using spatial analysis to identify and influence hard-to-reach customers, such as:

• Low-income

• Environmentally disadvantaged




• Low-income


Many utilities are considering new program designs that hinge on specific geographic energy usage patterns, such as:

• Demand-response

• Electric vehicles

• Non-wires solutions




• Low-income


Many utilities are considering new program designs that hinge on specific geographic energy usage patterns, such as:

• Demand-response

• Electric vehicles

• Non-wires solutions

M I C R O - T A R G E T I N G S P A T I A L P A T T E R N S

To better inform program design, utilities are combining program data with geographic data such as:

• Weather

• Traffic

• Demographics

S E G M E N T A T I O N

“ W H Y D O E S I T M A T T E R ? ”L A R G E S H I F T S I N G E O C O D I N G S E R V I C E S

O L D I N F O I N S T A N D -A L O N E S E R V I C E S

C H A N G E I N A P I R E Q U I R E M E N T S

C H A N G E S T O S E R V I C E C A P A B I L I T I E S

1 .

2.

3.





1 .

2.

3.https://www.esri.com/arcgis-blog/products/analytics/analytics/data-and-maps-for-arcgis-2015-is-now-available/

https://www.esri.com/arcgis-blog/products/analytics/analytics/data-and-maps-for-arcgis-2015-is-now-available/





1 .

2.

3.https://www.littlemissdata.com/blog/ggmap-updated

“Google has recently changed its API requirements… users are now required to provide an API key and enable billing.”

https://www.littlemissdata.com/blog/ggmap-updated





1 .

2.

3.https://multiplottr.com/

https://multiplottr.com/

“ W H Y D O E S I T M A T T E R ? ”W H A T A C C U R A C Y I S N E E D E D A N D A T W H A T C O S T ?

“ W H Y D O E S I T M A T T E R ? ”W H A T A C C U R A C Y I S N E E D E D A N D A T W H A T C O S T ?

https://community.tableau.com/thread/232540

https://community.tableau.com/thread/232540

A G E N D A



2. Introduction

3. Methods

4. Results

5. Conclusion

I N T R O D U C T I O NR E V I E W O F G E O C O D I N G S E R V I C E S A N D S I T E S

G E O C O D I N G S E R V I C EF R E E O R P A I D

S E R V I C EH A S A P I ?

Alteryx Free

ArcGIS Paid

BatchGeo Free and Paid

GeoLytics Paid

Google Free and Paid

Google Sheets Free and Paid

HERE Free and Paid

MapLarge Paid

OpenStreetMap Free

Texas A&M Geoservices Geocoding Free and Paid

TomTom Free and Paid

United States Census Bureau Geocoder API Free

I N T R O D U C T I O NC L I E N T A N D P R O G R A M

C L I E N T

AEP Ohio: Electric investor-owned utility serving nearly 1.5 million customers.


C L I E N T G E O G R A P H Y


Majority of AEP Ohio customers located in and around Columbus, OH.


G E O G R A P H Y D A T A

Majority of AEP Ohio customers located in and around Columbus, OH.

Examined the geolocational data of 151 AEP Ohio residential survey respondents who completed a Qualtrics web survey in January 2019.

Survey respondents part of residential rebate program evaluation.

C L I E N T


A G E N D A



2. Introduction

3. Methods

4. Results

5. Conclusion

M E T H O D SD A T A S E T S

R A W D A T A S E T

Raw dataset of survey respondent service addresses as received from AEP Ohio.


R A W D A T A S E T C L E A N D A T A S E T


Cleaned dataset of survey respondent service addresses using cleaning best-practices recommended by the Harvard School of Public Health (HSPH 2017).


R A W D A T A S E T C L E A N D A T A S E T


Cleaned dataset of survey respondent service addresses using cleaning best-practices recommended by the Harvard School of Public Health (HSPH 2017).

Differences in the geocoded locations of raw data and the cleaned data would give an indication of how effectively each service handles data quality issues.

H Y P O T H E S I S

M E T H O D SG E O C O D I N G S O U R C E S T E S T E D

U S E D G O O G L E A P I A S “ B A S E L I N E ”

P E R F O R M E D S P O T -C H E C K O N 2 0 R E C O R D S

C O M P A R E D B A S E L I N E T O 6 O T H E R S E R V I C E S

1 .

2.

3.





1 .

2.

3.

X





1 .

2.

3.

G E O C O D I N GS E R V I C E

G E O C O D I N G T Y P EG E O C O D E D

A SB A S E L I N E

G E O C O D E SP E R F O R M E D

O N B O T HR A W A N D

C L E A NA D D R E S S

D A T A

G E O C O D E SB A S E D O N

I P A D D R E S S

2013 Garmin

Map layer used as address locator tool in ArcMap

ArcGIS Online Online application

Census API API

Google API API

Google Sheets

Script within online application

OSM API API

QualtricsLocational data provided with survey responses

A G E N D A



2. Introduction

3. Methods

4. Results

5. Conclusion

R E S U L T SS U M M A R Y S T A T I S T I C S F R O M C O M P A R I S O N T O B A S E L I N E

In this particular experiment, found little difference in how the services geocoded the raw addresses and the cleaned addresses.

Likely speaks to good data practices on the part of AEP Ohio rather than the capabilities of the geocoding services.

R A W V S C L E A N




• Qualtrics was the outlier.

R A W V S C L E A N B A S E L I N E V S O T H E R S E R V I C E S


U N M A T C H E DG E O C O D I N G

R E C O R D S

A V E R A G E D I S T A N C EF R O M B A S E L I N EC O O R D I N A T E ( I N

M I L E S )

M E D I A N D I S T A N C EF R O M B A S E L I N EC O O R D I N A T E ( I N

M I L E S )

Google Sheets 0 < 0.01 < 0.01

Census API 10 0.05 0.02

2013 Garmin 0 0.22 0.02

OSM API 61 0.32 0.11

Qualtrics 0 95.30 6.15





• OSM and the Census did not find matches for all addresses.




R E C O R D S


M I L E S )


M I L E S )


Census API 10 0.05 0.02

2013 Garmin 0 0.22 0.02

OSM API 61 0.32 0.11






• OSM and the Census did not find matches for all addresses.

• Excluding Sheets, Census API had results closest to baseline.




R E C O R D S


M I L E S )


M I L E S )


Census API 10 0.05 0.02

2013 Garmin 0 0.22 0.02

OSM API 61 0.32 0.11


R E S U L T SF U R T H E R E X P L O R I N G D I F F E R E N C E S

Q U A L T R I C S : S U R V E Y R E S P O N D E N T L O C A T I O N

G E O C O D I N G S E R V I C E S V I S U A L C O M P A R I S O N

C O M P A R I N G B A S E L I N E T O A R C G I S O N L I N E

1 .

2.

3.

A G E N D A



2. Introduction

3. Methods

4. Results

5. Conclusion

C O N C L U S I O NT I P S A N D T R I C K S

Geocode a small amount of records (less than 100): may be easiest to use to Google Sheets.

S M A L L B A T C H



Geocode a larger amount of records (less than 40k): using the Google API may be free to use.*

S M A L L B A T C H L A R G E B A T C H

* Statement was true at the time the paper was written.


S M A L L B A T C H L A R G E B A T C H

Can perform a second set of geocoding on a subset of records using a different geocoding service.

Can check the subset of geocodes that are flagged as potentially less accurate based on the geocoders accuracy field.

I N C R E A S E A C C U R A C Y

* Statement was true at the time the paper was written.

Geocode a larger amount of records (less than 40k): using the Google API may be free to use.*


C O N C L U S I O NS U M M A R Y T A B L E


R E C O M M E N D E D U S E

C O S T F O R S M A L LN U M B E R O FG E O C O D E DL O C A T I O N S

C O S T F O R L A R G EN U M B E R O FG E O C O D E DL O C A T I O N S

A C C U R A C YI N C L U D E SA C C U R A C Y

M E T R I C

Google API Accurate geocoding of a small number of records; have help from a proficient coder

Free (less than 40,000

calls a month)

High(between $4 and

$5 per 1,000 calls)High Yes

Google Sheets Accurate geocoding of a small record batch.Free

(unclear when pay barrier starts)

Unknown(unclear when pay

barrier starts)High No

Census API

Fairly accurate geocoding of large number of records; unmatched records won’t pose an issue; have help from a proficient coder; pair with another service for unmatched addresses

Free Moderate –High Yes

2013 Garmin North America

Fairly accurate geocoding of large number of records; fast processing time; own Garmin map layers; have help from an ArcGIS user

Cost of acquiring Garmin map layer (cost unknown)

Moderate –High Yes

OSM API

Less accurate geocoding of large number of records; unmatched records won’t pose an issue; have help from a proficient coder; pair with another service for unmatched addresses

Free Low –Moderate No

Qualtrics Quick check that survey respondents are answering in roughly the predicted pattern

Free (no additional cost with Qualtrics service) Low No

ArcGIS Online Accurate geocoding; high price Approximately $4 for 1,000 geocodes High Yes

C O N C L U S I O NW H A T A C C U R A C Y I S N E E D E D A N D A T W H A T C O S T ?

C O N C L U S I O NW H A T A C C U R A C Y I S N E E D E D A N D A T W H A T C O S T ?

https://oehha.ca.gov/calenviroscreen/maps-data

https://oehha.ca.gov/calenviroscreen/maps-data

42

Questions?

Robert [email protected]

t. 206.388.0973

ALL OVER THE MAP: COMPARING THE ACCURACY OF …...Accurate geocoding of a small record batch. Free...

Documents

Transcript of ALL OVER THE MAP: COMPARING THE ACCURACY OF …...Accurate geocoding of a small record batch. Free...