This is a class project by:
Univeristy of Texas at Austin
School of Journalism
Data Visualization
Spring 2015
Instructor: Christian McDonald
We created a geocoded data set of all 9354 schools in Texas that can be merged with other data files from the Texas Education Agency for easy mapping. The files are available for free use.
The work was completed in April 2015.
- TEA has a file of these lat/longs that is probably a better source at this time: http://schoolsdata2-tea-texas.opendata.arcgis.com/, click on "Schools", then Download Spreadsheet.
- And Common Core has all schools in the US: https://nces.ed.gov/ccd/pubschuniv.asp
You can use the Download Zip button to just download all the files, or do a Save As on this file to get just the finished geocoded list, called texas-schools-geocoded.csv. It is 9354 rows of data. Feel free to fork this repo, update or improve the list and submit a pull request.
There are three sets of addresses in the file: The district, the school mailing address and the school site address. The Latitude and Longitude fields are keyed to the SchoolSiteStreetAddress.
The magic comes with the SchoolNumber field, which is a unique ID that matches other school-level data by the Texas Education Agency. You can database queries and data visualization tools to join data sets on that SchoolNumber to map accountability ratings and performance reports.
Things to know if you are using this data:
- Latitudes and Longitudes can only be as good as the address they start from. This will be particularly true for alternative schools, which can be ambiguous.
- Manually-found locations are based off Google Maps, and subject to any errors there. In many cases, the manual locations were verified through Street View photos.
- In some cases we just couldn't find the school. Those have blank Lat/Longs.
- The
CountyNameandCountyNumberfields are tied to the district, not the school, which is not always the same county asSchoolSiteStreetAddress. This is especially true with private schools.
texas-schools-geocoded.csvis a comma separated values file of geocoded schools.schoolsraw.csvis the origial file the project started from. It is called School and District File with Site Address, and was downloaded from the Texas Education Agency AskTED on 3/14/2015.csv_splitteris a python script used to split the raw file into pieces so each student in the class could have a batch of files to geocode. Someone might find that useful some day.
If a geocoding result was not AddressRangeInterpolation or ExactParcelCentroidPoint, the location was manually checked, and the GeocodeQualityType field changed to Manual. Other fields from the geocoding process have been removed for clarity.
We used Texas A&M Geoservices to batch geocode the files. It is a wonderful service that not only provides latitudes and longitudes for addresses, but gives a quality index of how the points were found. This is essential to be sure that an address was found at street point instead of a city or zip code centroid. Their service is free for up to 2500 addresses a month, or something like that. Check with them ... they are good people.
Geo-stuff provided by Texas A&M University GeoServices
The students followed this guide for the work. I checked and merged the individual files into a single data set.
If you have any corrections to offer, send them to christian.mcdonald@utexas.edu.
The following students from the University of Texas at Austin School of Journalism, Spring 2015 Data Visualization course contributed to this project:
- LUQMAN ADENIYI
- JOSEPH BAUCUM
- NICOLE COBLER
- TAYLOR ELLIS
- LAUREN FLORENCE
- TRACY FRYDBERG
- DANIEL GOODWIN
- SEBASTIAN HERRERA
- GABRIEL MACIAS
- WESLEY MARTIN
- RENEE MORENO
- ALEXIS SCHRUBBE
- PHILLIP TRACY
- BRYAN WINTER
- MANZHI WU
Also contributing:
- CHRISTIAN MCDONALD
- ANDRE MONTEIRO
- EDWARD TIMMS
Many thanks to all who helped. Enjoy.