Geocoding Tools

Tools for processing shapefiles and looking up coordinates by address or parcel ID.

Setup

1. Create and activate a virtual environment:

python -m venv env

source env/bin/activate # On Windows, use: env\Scripts\activate

2. Install required packages:

pip install pandas geopandas shapely

3. Download the Montgomery County GIS parcel shapefile from the Montgomery County GIS website and unzip it.

Usage

Process Shapefile

Convert a shapefile to CSV with centroid coordinates:

python process_shapefile.py

By default, this looks for out.shp and creates output.csv. You can modify the input/output paths in the script.

Look Up Coordinates

The lookup_coordinates script now supports two commands: lookup and process.

Looking up coordinates:

# Search by address (default)
python lookup_coordinates.py lookup "123 MAIN ST"

# Search by partial address
python lookup_coordinates.py lookup "MAIN"

# Search by parcel ID
python lookup_coordinates.py lookup "R72 12307 0032" --field parcel_id

# Use a different csv file
python lookup_coordinates.py lookup "MAIN" --csv path/to/coordinates.csv

Processing and joining data:

python lookup_coordinates.py process --csv output.csv --db parcels.db --output result.csv

Lookup Options:

• search_term: Address or parcel ID to search for
• --csv: Path to the CSV file (default: output.csv)
• --field: Field to search in ('address' or 'parcel_id', default: address)

Process Options:

• --csv: Input CSV file path
• --db: SQLite database path (default: parcels.db)
• --output: Output CSV path (default: result.csv)
• --id-field: Parcel ID field name (default: TAXPINNO)

Convert CSV to JSON

Convert the result CSV file to a JSON array:

python csv_to_json.py result.csv
# or specify custom output path
python csv_to_json.py result.csv --output data.json

The script will create a JSON file with the same name as the input CSV (but with .json extension) if no output path is specified.

Web Service

The geocoder can run as a REST API (FastAPI) that returns parcel coordinates by parcel ID or address.

1. Build the lookup index

The API reads from a compact, indexed SQLite database. The simplest way to build it needs no shapefile and no GIS libraries — it pages through the County's live parcel layer and asks for centroids:

python fetch_parcels.py          # County ArcGIS layer -> geocoder.db (~273k parcels, ~10 minutes)

The older route still works if you have the shapefile-derived CSV:

python build_index.py            # output.csv -> geocoder.db (~29MB)

2. Run locally

pip install -r requirements-web.txt
uvicorn app:app --reload

Interactive docs are served at http://localhost:8000/docs.

Endpoints

Endpoint	Description
GET /health	Liveness check; returns parcel count
GET /geocode?parcel_id=R72 12307 0032	Exact parcel ID lookup
GET /geocode?address=4060 DELPHOS	Partial, case-insensitive address lookup

Example:

curl "http://localhost:8000/geocode?address=4060%20DELPHOS"
# {"count":1,"results":[{"parcel_id":"R72 12307 0032","address":"4060 DELPHOS AVE",
#   "zip":"45402","latitude":39.763...,"longitude":-84.251...}]}

Provide exactly one of parcel_id or address. Address matches are substrings, so partial terms may return multiple results (capped at 50).

Deploy to Railway

The production service runs on Railway from this repo's Dockerfile, which downloads geocoder.db from the geocoder-data GitHub release. A monthly workflow (.github/workflows/geocoder-data.yml) rebuilds the index from the County layer, re-uploads it, and bumps LAST_BUILD so Railway redeploys. To refresh by hand:

python fetch_parcels.py
gh release upload geocoder-data geocoder.db --clobber

It is also the backend for the geocode tool in OpenDayton.

Deploy to Google Cloud Run

The geocoder.db index is baked into the container image (read-only), so the service is stateless and scales to zero when idle.

# Prereqs: gcloud CLI authenticated, a project selected, geocoder.db built.
./deploy.sh

To run the container locally:

docker build -t parcel-geocoder .
docker run -p 8080:8080 parcel-geocoder
curl "http://localhost:8080/health"

CSV File Structure

The script expects the following columns in the CSV file:

• LOC_NBR: Street number
• LOC_DIR: Street direction
• LOC_STREET: Street name
• LOC_SUFFIX: Street suffix
• TAXPINNO: Parcel ID
• latitude: Latitude coordinate
• longitude: Longitude coordinate

Known data issue: a few addresses point to the wrong place

About 190 parcels of 273,000 (0.07%) carry a street address that disagrees with where the county's own parcel polygon sits. R72 08705 0033 is recorded as 20 W Fourth St but lies roughly a mile west, out where the 1000 block is. Geocoding such an address returns the wrong location and nothing in the data says so.

These are the county's records and this service passes them through unchanged — inventing a corrected house number would be worse. Every build writes the current list to address_outliers.csv (report_address_outliers in fetch_parcels.py), which is the file to send the Auditor if you want them fixed at source.

The check compares only within a taxing district. Montgomery County has many municipalities and most have a Main St, so comparing by street name alone flags thousands of parcels that are simply in another town.

Notes

• Address searches are case-insensitive and support partial matches
• Multiple matches will be displayed if found
• Invalid geometries in the shapefile will be skipped during processing
• The process command in lookup_coordinates.py joins the CSV data with a SQLite database
• The csv_to_json.py script handles NaN values and provides proper JSON formatting