Skip to content

mlichter2/isolines

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

7 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

isolines

Street Networks Isolines and Isochrones

Create Isolines and Isochrones for street networks.

isolines is a Python library for creating street networks isolines (equal-distance) and isochrones (equal-time) polygons with just one line of code. It is built on top of Shapely, geopandas, NetworkX and osmnx.

isolines allows you to create isolines:

  • from OSM data without any data input or * use your own data: your own network and source locations

Main Features

  • Generation of isolines/isochroones for areas, segments, and point locations
  • Augmentation of the network for precise isoline/isochroone delineation
  • Based on a concave hull algorithm by default (Moreira, Adriano & Santos, Maribel. (2007) [1]) (for walking networks, otherwise convex hull is the default and concave hull is optional)
  • Source location can be either an address string to be geocoded using OSM Nominatim or a shapely geometry
  • Accepts various graph inputs: the GpdIsolimes class excepts edges GeoDataFrames and the NXIsoliner class accepts NetworkX graphs as input and the OsmIsolines class allows you tod downloads a street network from OpenStreetMap (OSM).

Other Features

  • Easy built-in visualization
  • Dynamically tweaking isolines/isochroones delineation by re-setting the default knn parameter that controls the concave hull heuristic
  • Dynamically adding isolines/isochroones to an existing instance
  • Extraction of the augmented output graph
  • Generation of nodes and edges GeoDataFrames

Examples

A repository with a wide variety of examples can be found here: https://github.com/mlichter2/isolines_examples

  • Create isolines/isochroones for complex geometries (polygons and linestrings) as well as simple point geometries

If you only want to output a GeoDataFrame of isolines/isochrones use the isolines function:

import isolines as il
df = il.isolines('Prospect Park, Brooklyn, NYC, USA', metric = 'time',values=[3, 6, 9])
df
geometry time
0 POLYGON ((-73.9715 40.6492,... 3
1 POLYGON ((-73.9708 40.6470,... 6
2 POLYGON ((-73.9704 40.6448... 9

However, if you which to explore your result with built-in visualization and be able to amend them and perform EDA use one of the following classes: GpdIsolimes, NXIsoliner, OsmIsolines

iso = il.OsmIsolines('Prospect Park, Brooklyn, NYC, USA', metric = 'time', values=np.arange(2.5, 22.5, 2.5), unit = 'ft', sample = 600)
iso.plot_isolines(figsize = (10, 10))

(for basic examples see also https://github.com/mlichter2/isolines_examples/blob/master/examples/01_basic_example.ipynb)

  • The isolines/isochroones boundaries are not confined to the existing network nodes. The network is amended to include new source and target nodes based on the input geometry and distances/times specified, so that large edges are cut in the respective location of new source/ target nodes to yield a more realistic isolines/isochroones boundary. the built-in visualization can can help explore and refine these boundaries.
iso = il.OsmIsolines('Bozeman High School, Bozeman, Montana, USA',values=[250, 500],sample = 200)
iso.plot_isolines(plot_nodes=True, plot_source_nodes=True,figsize = (10,10))

  • The output isolines/isochroones are based on a concave hull (rather than convex hull) heuristic by default resulting in a more realistically shaped and accurate isolines/isochroones. These can be tweaked to get even more refined delineation if needed
iso.change_isolines(knn = 15)
iso.plot_isolines(plot_nodes=True, plot_source_nodes=True,figsize = (10,10))

(examples for refining boundaries can be found here: https://github.com/mlichter2/isolines_examples/blob/master/examples/03_refining_isolines_concave_boundaries_and_smoothing.ipynb)

  • isolines lets you either download a network from OpenStreetMap (OSM) or use an edges geopandas GeoDataFrame or use a NetworkX graph In the example below, an edges GEoDataFrame from the US Census Tiger roads dataset is used
import geopandas as gpd
import pandas as pd
from shapely.geometry import LineString
df = gpd.GeoDataFrame.from_file('../data/tl_2019_36047_edges/tl_2019_36047_edges.shp')
# pre-process: add edges in the opposite direction 
df2 = df.copy()
df2['TNIDF'] = df['TNIDT'].copy()
df2['TNIDT'] = df['TNIDF'].copy()
# reverse the line geometry coordinate sequence
df2['geometry'] = df['geometry'].apply(lambda x: LineString(x.coords[::-1]))
df = pd.concat([df, df2]).reset_index(drop = True)
tiger = il.GpdIsolines('Prospect Park, Brooklyn, NYC, USA',
                            edges = df,
                            network_type = 'walk',
                            metric = 'time',
                            values=[3, 6, 8, 16, 20],
                            edge_idcol = 'TLID', 
                            fromcol = 'TNIDF',
                            tocol = 'TNIDT',
                            sample= 400,
                            knn = 50
                                 )

tiger.plot_isolines(figsize = (10, 10))

(examples for using different data sources: https://github.com/mlichter2/isolines_examples/blob/master/examples/06_using_different_input_sources.ipynb,

https://github.com/mlichter2/isolines_examples/blob/master/examples/07_example_using_US_Census_TIGER_dataset.ipynb)

  • Add isolines to an existing instance
from shapely.geometry import Point
isochrones = il.OsmIsolines(Point(179370.985,664422.488),
                                 network_type = 'walk',
                                 values=[500, 1500, 2500, 3500],
                                 crs = 2039,
                                 knn = 25)
isochrones.plot_isolines(figsize = (10, 10))

isochrones.add_isolines([1000, 2000, 3000, 4000])
isochrones.plot_isolines()

(examples: https://github.com/mlichter2/isolines_examples/blob/master/examples/02_adding_isolines_and_isochrones.ipynb)

  • Extract nodes and edges GeoDataFrames, NetworkX graphs
nodes = isochrones.get_nodes()
edges = isochrones.get_edges()
G = isochrones.get_graph()
  1. Moreira, Adriano & Santos, Maribel. (2007). Concave hull: A k-nearest neighbours approach for the computation of the region occupied by a set of points.. 61-68.