CCD Street Inventory Data Cleaning + CurbLR Feed

Phase 1: Generate Hourly Parking Table

Info

Additional data exploration info here.

Getting started

make sure ParkingZones.gdb is in data/raw

run:

export GEODATABASE=data/raw/ParkingZones.gdb
python -m ccd.clean $GEODATABASE

for a test examples (processing only the first 50 segments), add the --test flag:

python -m ccd.clean $GEODATABASE --test

Unit tests

To run all unit tests, run:

python -m unittest discover -s ccd/tests/ -p "test_*.py"

Phase 2: Generate CurbLR Dataset

Getting started

Requirements: Docker 18+

Move the CCD curbs geojson (e.g. CCD_curbs.geojson to the data directory)

Run the linear referencing script: ./scripts/linear_reference_all_radii.py --ccd_geoms=data/CCD_curbs.geojson --key=CCD_non-offset --max_radius=30

In the example above:

• data/CCD_curbs.geojson is the input CCD geometry file
• CCD_non-offset specifies the name of the directory within data where the linear referencing output will be stored.
• 30 is the maximum radius that the linear referencing tool will search.

If you would like for the linear reference lines to be offset from the street centerline, add the offset parameter:

./scripts/linear_reference_all_radii.py --ccd_geoms=data/CCD_curbs.geojson --key=CCD_non-offset --max_radius=30 --offset=5

Build the Docker image:

./docker/build

Move the hourly parking table (i.e. Parking_Regulations.csv) and the regulation lookup JSON file (i.e. reg_lookup.json) to the data directory.

The script to generate the CurbLR data will take the following inputs:

• segment_uri: Path to original CCD geometries (data/CCD_curbs.geojson)
• hour_table_uri: Path to the hourly parking table (data/Parking_Regulations.csv)
• linear_ref_dir: Directory with linear reference output that you created in the linear referencing step (data/CCD_non-offset/)
• reg_lookup_uri: Path to regulation lookup JSON (data/reg_lookup.json)
• output_key: 'Name of directory to be created for all output (CCD_CurbLR_non-offset)

Run the Docker container:

./docker/run

This will open a console within the container. Run the command to generate the entire CurbLR dataset:

./scripts/generate_curblr.py \
    --segment_uri=data/CCD_curbs.geojson \
    --hour_table_uri=data/Parking_Regulations.csv \
    --linear_ref_dir=data/CCD_non-offset/ \
    --reg_lookup_uri=data/reg_lookup.json \
    --output_key=CCD_CurbLR_non-offset

When the script completes, you will be able to find the results in data/CCD_CurbLR_non-offset/