def execute(trial=False, logging=True):
startTime = datetime.datetime.now()
if logging:
print("in make_graph.py")
# Set up the database connection.
client = dml.pymongo.MongoClient()
repo = client.repo
repo.authenticate('adsouza_bmroach_mcaloonj_mcsmocha',
'adsouza_bmroach_mcaloonj_mcsmocha')
points = repo[
'adsouza_bmroach_mcaloonj_mcsmocha.signal_placements'].find()
coords = points[0]['signal_placements']
df_geo = pd.DataFrame(coords, columns=['Lats', "Longs"])
# Turn a dataframe containing point data into a geojson formatted python dictionary
def df_to_geojson(df, properties, lat='Lats', lon='Longs'):
geo_json = {'type': 'FeatureCollection', 'features': []}
for _, row in df.iterrows():
feature = {
'type': 'Feature',
'properties': {},
'geometry': {
'type': 'Point',
'coordinates': []
}
}
feature['geometry']['coordinates'] = [row[lon], row[lat]]
for prop in properties:
feature['properties'][prop] = row[prop]
geo_json['features'].append(feature)
return geo_json
placements_geojson = df_to_geojson(df_geo, properties="")
with open('./adsouza_bmroach_mcaloonj_mcsmocha/placements.html',
'w') as output:
output.write(
geoleaflet.html(placements_geojson)) # Create visualization.
repo.logout()
endTime = datetime.datetime.now()
return {"start": startTime, "end": endTime}
def execute(trial=False, logging=True):
startTime = datetime.datetime.now()
#__________________________
#Parameters
mean_skew = 1.0
# ^ allows mean (for checking bottom 50th percent of distances) to be skewed, to allow in more or less.
# decreasing value decreases the mean, so fewer are allowed in
assert type(mean_skew) == float and mean_skew > 0
if trial:
keep_within_value = .5
else:
keep_within_value = 2
#End Parameters
#__________________________
if logging:
print("in fetch_nodes.py")
client = dml.pymongo.MongoClient()
repo = client.repo
repo.authenticate('adsouza_bmroach_mcaloonj_mcsmocha', 'adsouza_bmroach_mcaloonj_mcsmocha')
g = geoql.loads(requests.get('http://bostonopendata-boston.opendata.arcgis.com/datasets/cfd1740c2e4b49389f47a9ce2dd236cc_8.geojson').text, encoding="latin-1")
g = g.keep_within_radius((42.3551, -71.0656), keep_within_value, 'miles')
g = g.node_edge_graph()
g.dump(open('example_extract.geojson', 'w'))
f = open('example_extract.geojson', 'r')
f = f.read()
j = json.loads(f)["features"]
#Creates dictionary with only nodes of the graph
nodes= [[obj['coordinates'][1],obj['coordinates'][0]] for obj in j if (obj['type'] == "Point")]
#print (len(nodes))
open('leaflet.html', 'w').write(geoleaflet.html(g)) # Create visualization.
clusters = repo['adsouza_bmroach_mcaloonj_mcsmocha.accident_clusters'].find_one()
clusters = clusters["accident_clusters"]
#Insert clusters into cKDTree
cluster_tree = cKDTree(clusters)
distances = []
#For each node, get distance of nearest cluster
for i in range(len(nodes)):
#find the k nearest neighbors
dist, idx = cluster_tree.query(nodes[i], k=1, p=2) #p=2 means euclidean distance
distances.append((dist,idx))
#Get average distance to closest cluster
mean = (sum([x[0] for x in distances])/len(distances)) * mean_skew
#Filter out nodes that have distance to nearest cluster that is less than the mean
filtered_nodes = [nodes[x[1]] for x in distances if x[0] >= mean]
#Insert into repo {"nodes": [[lat,long], [lat,long]..............]}
tmp = dict()
tmp['nodes'] = filtered_nodes
filtered_nodes = tmp
repo.dropCollection("nodes")
repo.createCollection("nodes")
repo['adsouza_bmroach_mcaloonj_mcsmocha.nodes'].insert(filtered_nodes)
repo['adsouza_bmroach_mcaloonj_mcsmocha.nodes'].metadata({'complete':True})
repo.logout()
endTime = datetime.datetime.now()
return {"start":startTime, "end":endTime}
import geojson
from geoql import geoql
import geoleaflet
import requests
#url = 'https://raw.githubusercontent.com/Data-Mechanics/geoql/master/examples/'
# Boston ZIP Codes regions.
#z = geoql.loads(requests.get(url + 'example_zips.geojson').text, encoding="latin-1")
# Extract of street data.
#g = geoql.loads(requests.get(url + 'example_extract.geojson').text, encoding="latin-1")
g = geoql.loads(requests.get(
'http://bostonopendata-boston.opendata.arcgis.com/datasets/cfd1740c2e4b49389f47a9ce2dd236cc_8.geojson'
).text,
encoding="latin-1")
# g = g.properties_null_remove()\
# .tags_parse_str_to_dict()\
# .keep_by_property({"highway": {"$in": ["residential", "secondary", "tertiary"]}})
g = g.keep_within_radius((42.3551, -71.0656), 2,
'miles') # 0.75 miles from Boston Common.
#g = g.keep_that_intersect(z) # Only those entries found in a Boston ZIP Code regions.
g = g.node_edge_graph() # Converted into a graph with nodes and edges.
g.dump(open('example_extract.geojson', 'w'))
open('leaflet.html', 'w').write(geoleaflet.html(g)) # Create visualization.
for f in tqdm(students.features,
desc='Updating student data with bus assignments'):
coords = f.geometry.coordinates
f['properties']['bus_id'] = school_stop_to_bus[(tuple(coords[2]),
tuple(coords[1]))]
open(file_students, 'w').write(geojson.dumps(students, indent=2))
return routes
if __name__ == "__main__":
grid = Grid('input/segments-prepared.geojson')
buses = json.load(open('output/buses.json', 'r'))
students = geojson.load(open('output/students.geojson', 'r'))
stops = stops_to_dict('output/stops.json')
routes = school_stops_to_routes(grid,
'output/students.geojson',
school_to_stops(stops),
buses,
max_dist_miles=20,
max_stops=30)
open('output/routes.geojson', 'w').write(
geojson.dumps(
geojson.FeatureCollection(
[f for r in routes for f in r.features()])))
open('output/routes.html', 'w').write(
geoleaflet.html(
geojson.FeatureCollection(
[f for r in routes for f in r.features()])))
## eof