CS_Spatial_Modules

Create spatial indicators and mobility simulations for CityScope.

OpenCity

A tool for creating simple mobility simulations and vizualisations with minimal data inputs.

Data

The data required are:

• a shapefile of geometries in the study region
• an O-D matrix, optionally stratified by demographic attributes and/or industry. eg:

Home GEOID	Work GEOID	N_total	N_income_low	...	N_NAICS_44-45	...
1	1	30	15	...	5	...
1	2	40	17	...	9	...
...	...	...	...	...	...	...

For USA-based projects the data above can be easily obtained from public data sources using the tools in this repository.

Usage

Build geometry for state

Specify

• the state where the simulation will take place
• the geometry type ('block_group' (default) or 'block')
• the year to use for the geometry and commuting data
• Optionally: a center coordinate and radius to restrict the analysis to

fips=25
year=2017
geom_type='block'
centre= {'lat': 42.352927, 'lon': -71.059435}
centre_x_y=[centre['lon'], centre['lat']]

model_area_radius=5000
sim_area_radius=2000

state=OpenCity.US_State(state_fips=fips, year=year, 
#                         geom_type=geom_type
                       )
state.get_geometry()

state.subset_geom_by_distance(centre_x_y, model_area_radius, 'model_area')
state.subset_geom_by_distance(centre_x_y, sim_area_radius, 'sim_area')

The state geometry can be accessed as a geopandas GeoDataFrame

all_zones=state.return_geometry()
sim_zones=state.return_geometry('sim_area')
model_zones=state.return_geometry('model_area')

Get commuting data for state from LEHD

state.get_lodes_data()

The commuting data can be used to build a simulated population

simpop_df=state.lodes_to_pop_table(model_subset_name='model_area',
                                  sim_subset_name='sim_area')

Build mobility system for same area

Get the road network(s). These are used to create pandana network objects.

import pandana
import osmnet

networks={}

bbox=state.get_bounds(subset_name='model_area')
drive_nodes_df,drive_edges_df=osmnet.load.network_from_bbox(lat_min=bbox[1], lng_min=bbox[0], lat_max=bbox[3], 
                          lng_max=bbox[2], bbox=None, network_type='drive', 
                          two_way=True, timeout=180, 
                          custom_osm_filter=None)
drive_edges_df['travel_time']=drive_edges_df['distance']*50000/3600

drive_net=pandana.Network(drive_nodes_df["x"], drive_nodes_df["y"], drive_edges_df["from"], drive_edges_df["to"],
                 drive_edges_df[["distance", "travel_time"]])

networks['drive']=OpenCity.PdnaNetwork(drive_net)

Define modes of transportation

drive_dict={
    'target_network_id': 'drive',
    'travel_time_metric': 'travel_time'}


modes={'drive': OpenCity.Mode(drive_dict)}

Create mobility system using the pandana network(s) and mode definition(s)

mob_sys=MobilitySystem(modes=modes,
                      networks=networks)

Create the Simulation Model

In order to reduct computational burden, the population can be further restricted to a maximum by using the "sample_N" parameter

sim=OpenCity.Simulation(simpop_df, mob_sys, model_zones)
simpop_df=sim.get_simpop_subset(simpop_df, sample_N=1000)

Simulate trips and trajectories

all_trips_df=sim.create_trip_table(simpop_df)
all_trips_df=sim.mode_chooser(all_trips_df)
route_table=sim.get_routes_table(all_trips_df)

Visualising Outputs

Get resulting trips as a GeoDataFrame of 'LineString's

route_gdf=sim.route_table_to_geo(route_table)

Get resulting trips as a geojsoon compatible with the kepler.gl Trips Layer. 'start_day_time_stamp' should be the timestamp in epoch seconds format at midnight at the beginning of the day being simulated.

start_day_time_stamp=
geo_dict=sim.route_gdf_to_trips_geojson(route_gdf, start_day_time_stamp)

CityScope Indicators

A set of modules which generate interactive indicators for CityScope project. Communication with cityIO is handled via the cs-brix library.

Data

The data required are:

• a shapefile of zone geometries in the study region. This should include columns indicating (i) counts of people living in each area by demographic categories and (ii) counts of people employed in each area by demographic categories.

GEOID	res_income_low	res_income_med	...	emp_total	emp_income_low	...	geometry
1	1000	500	300	...	600	...	POLYGON(...)
2	1200	600	200	...	550	...	POLYGON(...)
...	...	...	...	...	...	...	POLYGON(...)

• the GEOGRID of the CityScope table (can be created using CityScopeJS)
• the simulated population (can be created from an O-D matrix using OpenCity as outlined above)

Usage

import CS_Indicators as CS from brix import Indicator, Handler import geopandas as gpd import pandas as pd

geogrid=gpd.read_file(SAVED_GEOGRID_LOCATION) zones=gpd.read_file(SAVED_ZONES_LOCATION) simpop_df=pd.read_csv(SAVED_SIMPOP_LOCATION)

table_name='cs_course_volpe'

H=Handler(table_name=table_name) H.reset_geogrid_data() # Optional

d=CS.Density_Indicator(zones=zones) p=CS.Proximity_Indicator(zones=zones, geogrid=geogrid) m=CS.Mobility_indicator(zones, geogrid, table_name, simpop_df) H.add_indicators([ d, p, m ])

H.listen()