make_gtfsを使って仮のGTFS(時刻表)を生成する

README.md

make_gtfsを使って仮のGTFS(時刻表)を生成する (山陽電鉄の一部)

実際の発車時刻とは異なるが、1時間あたりの本数はなるべく合わせた時刻表。

駅の緯度経度や線路(ルート)形状の情報は、 国土数値情報 鉄道データを利用。

指定したbounding box内の駅のみを対象とする。

shapes.geojson

uv run python withinbbox.py 134.963436,34.623885,135.170288,34.738791 N02-24_RailroadSection.shp
grep -e '本線.*山陽電気鉄道' N02-24_RailroadSection.134.963436,34.623885,135.170288,34.738791.geojson \
| sed -e '1i {"type":"FeatureCollection","features":[' -e '$s/,$/]}/' >sanden.geojson
uv run python linemerge.py sanden.geojson
sed -e '1i {"type":"Feature","properties":{"shape_id":"shp1"},"geometry":' -e '$a }' sanden.linemerge.geojson >shapes.geojson

1本のLineStringにマージしてshapes.geojsonを作成。 (linemerge.pyは https://gist.github.com/deton/3508b7a291b81d1f9bbedad8ea26dccf と同じ)

service_windows.csvとfrequencies.csv

NAVITIMEで駅の時刻表を見て(列車種別で絞り込みしつつ)、 各曜日・各時間帯・上り/下りの本数を調べて記入。

1時間あたりの本数が同じ時間が続く場合は1行にまとめられる。

(とりあえず、列車種別:普通、平日を対象に手作業)

stops.csv

駅のgeometryとして、midpointの緯度経度に変換

uv run python withinbbox.py 134.963436,34.623885,135.170288,34.738791 N02-24_Station.shp
uv run python midpoint.py N02-24_Station.134.963436,34.623885,135.170288,34.738791.geojson

とりあえず、重複するstop_idや駅名は手で削除。

grep -e '本線.*山陽電気鉄道' N02-24_Station.134.963436,34.623885,135.170288,34.738791.midpoint.geojson | sed -e 's/,$//' > stops.ndjson
vim stops.ndjson

(参考:同名駅のグループに属すものを検索。 緯度経度が同じなら、N02_005gの値をstop_idとして使う。)

grep -v 'N02_005c": "\([^"]*\)", "N02_005g": "\1"' N02-24_Station.134.963436,34.623885,135.170288,34.738791.midpoint.geojson

echo 'stop_id,stop_name,stop_lon,stop_lat,zone_id' > stops.csv
jq -r '[.properties.N02_005c,.properties.N02_005,.geometry.coordinates[0],.geometry.coordinates[1],.properties.N02_005c]|@csv' stops.ndjson >>stops.csv

make_gtfsを実行

uv run make_gtfs data/sanden_local output/sanden_local.zip

運賃ファイルfare_rules.txt, fare_attributes.txtを生成

運賃表PDFを見て、下記の内容のfare.csvを作成。

price,min_km,max_km
170,0,2
200,2,4
250,4,7
320,7,10
390,10,13
470,13,17
540,17,21
600,21,25

uv run python make_gtfs_fare.py output/sanden_local.zip fare.csv

生成されるfare_rules.txtとfare_attributes.txtをGTFS zipに含める。

TimestampedGeoJsonに変換して表示してみる

https://gist.github.com/deton/8c4bd41faa4060dfa4a68066a5403be3#file-gtfsshape2timestampedgeojson-ipynb

https://deton.github.io/HeatMapWithTime/IconPositionWithTime.html?tripWidth=6&iconScale=0.7&zoom=12&pointRadius=2&lineWidth=1&latlon=34.65959,135.067874&jsonurl=https://gist.githubusercontent.com/deton/ac80db5e4b464169a764b524d3377c96/raw/7400f6ae9b1e758ffa5da9fb9d42c09cade52385/zzGtfsTimestamped.geojson

centroid.py

import sys
import geopandas as gpd
# geojson from https://nlftp.mlit.go.jp/ksj/gml/datalist/KsjTmplt-N02-2024.html
# 駅のgeometryとして、重心Pointの緯度経度に変換
infile = sys.argv[1]
outfile = '.'.join(infile.split('.')[:-1] + ['centroid', 'geojson'])
# TODO: 場所に応じた適切なCRSを使う
rails = gpd.read_file(infile, encoding="utf-8").to_crs('EPSG:6690')
rails['centroid'] = rails['geometry'].centroid.to_crs('EPSG:6668')
rails.drop(columns=['geometry'], inplace=True)
rails.to_file(outfile, driver='GeoJSON')

fare.csv

frequencies.csv

make_gtfs_fare.py

# make GTFS fare_attributes.txt and fare_rules.txt using following fare.csv.
# 
# ```csv
# price,min_km,max_km
# 170,0,2
# 200,2,4
# 250,4,7
# 320,7,10
# 390,10,13
# 470,13,17
# 540,17,21
# 600,21,25
# ```
import argparse
import gtfs_kit as gk
import pandas as pd

parser = argparse.ArgumentParser()
parser.add_argument("gtfspath", help="GTFS path or URL")
parser.add_argument("farecsv", help="fare.csv file with columns: price,min_km,max_km")
parser.add_argument("--currency_type", default="JPY")
parser.add_argument("--payment_method", type=int, choices=[0, 1], default=1)
parser.add_argument("--transfers", type=int, choices=[0, 1, 2], default=0)  # TODO: support empty
args = parser.parse_args()

CURRENCY_TYPE = args.currency_type
PAYMENT_METHOD = args.payment_method
TRANSFERS = args.transfers

feed = gk.read_feed(args.gtfspath, dist_units='m')
feed = feed.append_dist_to_stop_times()

target_triproute = feed.compute_trip_stats().groupby('route_id').first().reset_index()[['route_id','trip_id']].to_dict(orient='list')

week = feed.get_first_week()
date = week[0]  # Monday

dfstops = feed.get_stops(date)
dfst = feed.get_stop_times(date)
dfst = dfst.merge(dfstops)
tripsdict = dfst.groupby('trip_id').apply(lambda x: x.to_dict(orient='records'), include_groups=False).to_dict()

dffare = pd.read_csv(args.farecsv)
dffare['fare_id'] = 'f' + dffare['price'].map(str)

dffare_attr = dffare[['fare_id','price']].copy()
dffare_attr['currency_type'] = CURRENCY_TYPE
dffare_attr['payment_method'] = PAYMENT_METHOD
dffare_attr['transfers'] = TRANSFERS
dffare_attr.to_csv('fare_attributes.txt', index=False)

dffare_rules = pd.DataFrame()
for trip_id, stop_times_list in tripsdict.items():
    if trip_id not in target_triproute['trip_id']:
        continue
    route_id = target_triproute['route_id'][target_triproute['trip_id'].index(trip_id)]
    #stop_times_list.sort(key=lambda x: x['arrival_time'])
    for i in range(1, len(stop_times_list)):
        distoffset = stop_times_list[i-1]['shape_dist_traveled']
        origzoneid = stop_times_list[i-1]['zone_id']
        for st in stop_times_list[i:]:
            dist = st['shape_dist_traveled']
            distkm = (dist - distoffset) / 1000.0
            df = dffare[(distkm >= dffare['min_km']) & (distkm < dffare['max_km'])].copy()
            df['route_id'] = route_id
            df['origin_id'] = origzoneid
            df['destination_id'] = st['zone_id']
            #df['orig_stop'] = stop_times_list[i-1]['stop_name']
            #df['dest_stop'] = st['stop_name']
            dffare_rules = pd.concat([dffare_rules, df.drop(columns=['min_km', 'max_km', 'price'])])
dffare_rules.to_csv('fare_rules.txt', index=False)

meta.csv

midpoint.py

import sys
import geopandas as gpd
# geojson from https://nlftp.mlit.go.jp/ksj/gml/datalist/KsjTmplt-N02-2024.html
# 駅のgeometryとして、midpointの緯度経度に変換

EPSG = "EPSG:6668"

infile = sys.argv[1]
outfile = '.'.join(infile.split('.')[:-1] + ['midpoint', 'geojson'])

rails = gpd.read_file(infile, encoding="utf-8")
utmcrs = rails.estimate_utm_crs()
railsutm = rails.to_crs(utmcrs)

railsutm['midpoint'] = railsutm['geometry'].interpolate(0.5, normalized=True).to_crs(EPSG)
railsutm.to_crs(EPSG).drop(columns=['geometry']).to_file(outfile, driver='GeoJSON')

service_windows.csv

shapes.geojson

stops.csv

stops.ndjson

{ "type": "Feature", "properties": { "N02_001": "12", "N02_002": "4", "N02_003": "本線", "N02_004": "山陽電気鉄道", "N02_005": "板宿", "N02_005c": "007372", "N02_005g": "007372" }, "geometry": { "type": "Point", "coordinates": [ 135.133288242230975, 34.660156819922143 ] } }
{ "type": "Feature", "properties": { "N02_001": "12", "N02_002": "4", "N02_003": "本線", "N02_004": "山陽電気鉄道", "N02_005": "西代", "N02_005c": "007362", "N02_005g": "007361" }, "geometry": { "type": "Point", "coordinates": [ 135.143989998451048, 34.662400000769487 ] } }
{ "type": "Feature", "properties": { "N02_001": "12", "N02_002": "4", "N02_003": "本線", "N02_004": "山陽電気鉄道", "N02_005": "西新町", "N02_005c": "007426", "N02_005g": "007426" }, "geometry": { "type": "Point", "coordinates": [ 134.980804999752081, 34.64957500057676 ] } }
{ "type": "Feature", "properties": { "N02_001": "12", "N02_002": "4", "N02_003": "本線", "N02_004": "山陽電気鉄道", "N02_005": "須磨浦公園", "N02_005c": "007480", "N02_005g": "007480" }, "geometry": { "type": "Point", "coordinates": [ 135.100219998560675, 34.637935000991746 ] } }
{ "type": "Feature", "properties": { "N02_001": "12", "N02_002": "4", "N02_003": "本線", "N02_004": "山陽電気鉄道", "N02_005": "月見山", "N02_005c": "007420", "N02_005g": "007420" }, "geometry": { "type": "Point", "coordinates": [ 135.122118330359285, 34.650186405397477 ] } }
{ "type": "Feature", "properties": { "N02_001": "12", "N02_002": "4", "N02_003": "本線", "N02_004": "山陽電気鉄道", "N02_005": "山陽垂水", "N02_005c": "007512", "N02_005g": "007512" }, "geometry": { "type": "Point", "coordinates": [ 135.053739999694017, 34.629335001332265 ] } }
{ "type": "Feature", "properties": { "N02_001": "12", "N02_002": "4", "N02_003": "本線", "N02_004": "山陽電気鉄道", "N02_005": "大蔵谷", "N02_005c": "007444", "N02_005g": "007444" }, "geometry": { "type": "Point", "coordinates": [ 135.008405000672781, 34.646585000707546 ] } }
{ "type": "Feature", "properties": { "N02_001": "12", "N02_002": "4", "N02_003": "本線", "N02_004": "山陽電気鉄道", "N02_005": "人丸前", "N02_005c": "007437", "N02_005g": "007437" }, "geometry": { "type": "Point", "coordinates": [ 135.002600000284019, 34.647545000758335 ] } }
{ "type": "Feature", "properties": { "N02_001": "12", "N02_002": "4", "N02_003": "本線", "N02_004": "山陽電気鉄道", "N02_005": "舞子公園", "N02_005c": "007492", "N02_005g": "007492" }, "geometry": { "type": "Point", "coordinates": [ 135.03385500044638, 34.634225001004651 ] } }
{ "type": "Feature", "properties": { "N02_001": "12", "N02_002": "4", "N02_003": "本線", "N02_004": "山陽電気鉄道", "N02_005": "滝の茶屋", "N02_005c": "007504", "N02_005g": "007504" }, "geometry": { "type": "Point", "coordinates": [ 135.072919999421572, 34.631235000724921 ] } }
{ "type": "Feature", "properties": { "N02_001": "12", "N02_002": "4", "N02_003": "本線", "N02_004": "山陽電気鉄道", "N02_005": "東須磨", "N02_005c": "007400", "N02_005g": "007400" }, "geometry": { "type": "Point", "coordinates": [ 135.1274849984909, 34.65521500058918 ] } }
{ "type": "Feature", "properties": { "N02_001": "12", "N02_002": "4", "N02_003": "本線", "N02_004": "山陽電気鉄道", "N02_005": "山陽明石", "N02_005c": "007429", "N02_005g": "007429" }, "geometry": { "type": "Point", "coordinates": [ 134.992949761341464, 34.648798424835398 ] } }
{ "type": "Feature", "properties": { "N02_001": "12", "N02_002": "4", "N02_003": "本線", "N02_004": "山陽電気鉄道", "N02_005": "須磨寺", "N02_005c": "007449", "N02_005g": "007449" }, "geometry": { "type": "Point", "coordinates": [ 135.116239999376177, 34.646025000127366 ] } }
{ "type": "Feature", "properties": { "N02_001": "12", "N02_002": "4", "N02_003": "本線", "N02_004": "山陽電気鉄道", "N02_005": "霞ヶ丘", "N02_005c": "007508", "N02_005g": "007508" }, "geometry": { "type": "Point", "coordinates": [ 135.042720001598951, 34.630325000696565 ] } }
{ "type": "Feature", "properties": { "N02_001": "12", "N02_002": "4", "N02_003": "本線", "N02_004": "山陽電気鉄道", "N02_005": "山陽須磨", "N02_005c": "007457", "N02_005g": "007457" }, "geometry": { "type": "Point", "coordinates": [ 135.112314999141461, 34.64357500160429 ] } }
{ "type": "Feature", "properties": { "N02_001": "12", "N02_002": "4", "N02_003": "本線", "N02_004": "山陽電気鉄道", "N02_005": "東垂水", "N02_005c": "007511", "N02_005g": "007511" }, "geometry": { "type": "Point", "coordinates": [ 135.063589999696802, 34.629375000624158 ] } }
{ "type": "Feature", "properties": { "N02_001": "12", "N02_002": "4", "N02_003": "本線", "N02_004": "山陽電気鉄道", "N02_005": "林崎松江海岸", "N02_005c": "007415", "N02_005g": "007415" }, "geometry": { "type": "Point", "coordinates": [ 134.965080000648328, 34.652280000372073 ] } }
{ "type": "Feature", "properties": { "N02_001": "12", "N02_002": "4", "N02_003": "本線", "N02_004": "山陽電気鉄道", "N02_005": "山陽塩屋", "N02_005c": "007496", "N02_005g": "007496" }, "geometry": { "type": "Point", "coordinates": [ 135.082039999504104, 34.633615000533077 ] } }
{ "type": "Feature", "properties": { "N02_001": "12", "N02_002": "4", "N02_003": "本線", "N02_004": "山陽電気鉄道", "N02_005": "西舞子", "N02_005c": "007474", "N02_005g": "007474" }, "geometry": { "type": "Point", "coordinates": [ 135.02813000127972, 34.638700000694072 ] } }

withinbbox.py

import sys
import geopandas as gpd
from shapely.geometry import Polygon

# Usage: python3 withinbbox.py 134.963436,34.623885,135.170288,34.738791 railstation.centroid.geojson
bb = [float(x) for x in sys.argv[1].split(',')]
infile = sys.argv[2]
outfile = '.'.join(infile.split('.')[:-1] + [sys.argv[1], 'geojson'])

coords = ((bb[0], bb[1]),
          (bb[0], bb[3]),
          (bb[2], bb[3]),
          (bb[2], bb[1]))
bbox = Polygon(coords)

gdf = gpd.read_file(infile, encoding='utf-8')
gdf[gdf.within(bbox)].to_file(outfile, driver='GeoJSON')

zzGtfsTimestamped.geojson