SDV (Synthetic Data Vault) trial using GBFS data

00README.md

sdv-trial-gbfs

GaussianCopulaSynthesizer

• GaussianCopulaSynthesizer

CTGANSynthesizer

• CTGANSynthesizer

TVAESynthesizer

• TVAESynthesizer

License

• *.py files are distributed under the terms of the MIT license.
• docomo-cycle.csv and hellocycling.csv:
  • These sample data use the following copyrighted material with modifications.
    • DOCOMO BIKESHARE, INC. / Association for Open Data of Public Transportation, Bikeshare information of DOCOMO BIKESHARE, INC., CC BY 4.0.
    • OpenStreet Corp. / Association for Open Data of Public Transportation, Bikeshare information of OpenStreet, CC BY 4.0, ODC BY 1.0, ODbL 1.0.

csv2hist.bat

python3 csv2hist.py sdv-gc-d.csv docomo-cycle >sdv-gc.ndjson
python3 csv2hist.py sdv-gc-h.csv hellocycling >>sdv-gc.ndjson
python3 csv2hist.py sdv-ctgan-d.csv docomo-cycle >sdv-ctgan.ndjson
python3 csv2hist.py sdv-ctgan-h.csv hellocycling >>sdv-ctgan.ndjson
python3 csv2hist.py sdv-tvae-d.csv docomo-cycle >sdv-tvae.ndjson
python3 csv2hist.py sdv-tvae-h.csv hellocycling >>sdv-tvae.ndjson

csv2hist.py

import json
import sys

fname = sys.argv[1]
systemId = sys.argv[2]

statuslist = []
stinfo = {}
with open(fname) as f:
    for line in f:
        line = line.rstrip('\r\n')
        stationId, lat, lon, capacity = line.split(',')
        if stationId == 'station_id':  # header line
            continue
        stinfo[stationId] = [float(lat), float(lon)]
        statuslist.append([stationId, int(capacity)])

outdic = {
    'ts': '2024-08-15T00:02',
    'system_id': systemId,
    'status': statuslist,
    'stations': stinfo,
}
print(json.dumps(outdic, ensure_ascii=False, separators=(',', ':')))

docomo-cycle.csv.gz

gbfs2capacitycsv.py

# Usage: python3 gbfs2capacitycsv.py 15/0002hellocycling_station_information.json
import json
import sys


print('station_id,lat,lon,capacity')
for stfname in sys.argv[1:]:
    with open(stfname) as f:
        info = json.load(f)
    for s in info['data']['stations']:
        capacity = s.get('capacity', s.get('vehicle_capacity'))
        print(f"{s['station_id']},{s['lat']},{s['lon']},{capacity}")

hellocycling.csv.gz

sdv-ctgan-d.csv.gz

sdv-ctgan-h.csv.gz

sdv-ctgan.geojson

sdv-gc-d.csv.gz

sdv-gc-h.csv.gz

sdv-gc.geojson

sdv-trial-gbfs.ipynb

sdv-tvae-d.csv.gz

sdv-tvae-h.csv.gz

sdv-tvae.geojson