intro.md

layout
@src/layouts/PageLayout.astro

Getting Started

cityseer is a collection of computational tools for fine-grained street-network and land-use analysis, useful for assessing the morphological precursors to vibrant neighbourhoods.

cityseer is underpinned by network-based methods that have been developed from the ground-up for micro-morphological urban analysis at the pedestrian scale, with the intention of providing contextually specific metrics for any given streetfront location. Importantly, cityseer computes metrics directly over the street network and offers distance-weighted variants. The combination of these strategies makes cityseer more contextually sensitive than methods otherwise based on crow-flies aggregation methods that do not take the network structure and its affect on pedestrian walking distances into account.

The use of python facilitates interaction with popular computational tools for network manipulation (e.g. networkX), geospatial data processing (e.g. shapely, etc.), Open Street Map workflows with OSMnx, and interaction with the numpy, geopandas (and momepy) stack of packages. The underlying algorithms are parallelised and implemented in rust so that the methods can be scaled to large networks. In-out convenience methods are provided for interfacing with networkX and graph cleaning tools aid the incorporation of messier network representations such as those derived from Open Street Map.

The github repository is available at github.com/benchmark-urbanism/cityseer-api. Please cite the associated paper when using this package. Associated papers introducing the package and demonstrating the forms of analysis that can be done with it are available at arXiv.

Installation

cityseer is a python package that can be installed with pip:

pip install cityseer

Code tests are run against Python versions 3.10 - 3.12.

Notebooks

The getting started guide on this page, and a growing collection of other examples, is available as an Jupyter Notebooks which can be accessed via the Examples page. The examples include workflows showing how to run network centralities and land-use accessibility analysis for some real-world situations.

Old versions

For documentations of older versions of cityseer, please refer to the docstrings which are directly embedded in the code for the respective version:

• Documentation for v1.x
• Documentation for v2.x
• Documentation for v3.x

Quickstart

cityseer revolves around networks (graphs). If you're comfortable with numpy and abstract data handling, then the underlying data structures can be created and manipulated directly. However, it is generally more convenient to sketch the graph using NetworkX and to let cityseer take care of initialising and converting the graph for you.

# any networkX MultiGraph with 'x' and 'y' node attributes will do
# here we'll use the cityseer mock module to generate an example networkX graph
import networkx as nx
from cityseer.tools import mock, graphs, plot, io

G = mock.mock_graph()
print(G)
# let's plot the network
plot.plot_nx(G, labels=True, node_size=80, dpi=200, figsize=(4, 4))

An example graph.

Graph Preparation

The tools.graphs module contains a collection of convenience functions for the preparation and conversion of networkX MultiGraphs, i.e. undirected graphs allowing for parallel edges. The tools.graphs module is designed to work with raw shapely Linestring geometries that have been assigned to the graph's edge (link) geom attributes. The benefit to this approach is that the geometry of the network is decoupled from the topology: the topology is consequently free from distortions which would otherwise confound centrality and other metrics.

There are generally two scenarios when creating a street network graph:

1. In the ideal case, if you have access to a high-quality street network dataset -- which keeps the topology of the network separate from the geometry of the streets -- then you would construct the network based on the topology while assigning the roadway geometries to the respective edges spanning the nodes. OS Open Roads is a good example of this type of dataset. Assigning the geometries to an edge involves A) casting the geometry to a shapely LineString, and B) assigning this geometry to the respective edge by adding the LineString geometry as a geom attribute. e.g. G.add_edge(start_node, end_node, geom=a_linestring_geom).

2. In reality, most data-sources are not this refined and will represent roadway geometries by adding additional nodes to the network. For a variety of reasons, this is not ideal and you may want to follow the Graph Cleaning guide; in these cases, the graphs.nx_simple_geoms method can be used to generate the street geometries, after which several methods can be applied to clean and prepare the graph. For example, nx_wgs_to_utm aids coordinate conversions; nx_remove_dangling_nodes removes remove roadway stubs, nx_remove_filler_nodes strips-out filler nodes, and nx_consolidate_nodes assists in cleaning-up the network.

3. A third approach uses only minimal graph cleaning and compensates for messy network representations through algorithmic corrections. This is demonstrated in Graph Corrections example notebook on the Examples page.

Example

Here, we'll walk through a high-level overview showing how to use cityseer. You can provide your own shapely geometries if available; else, you can auto-infer simple geometries from the start to end node of each network edge, which works well for graphs where nodes have been used to inscribe roadway geometries.

# use nx_simple_geoms to infer geoms for your edges
G = graphs.nx_simple_geoms(G)
plot.plot_nx(G, labels=True, node_size=80, plot_geoms=True, dpi=200, figsize=(4, 4))

A graph with inferred geometries. In this case the geometries are all exactly straight.

We have now inferred geometries for each edge, meaning that each edge now has an associated LineString geometry. Any further manipulation of the graph using the cityseer.graph module will retain and further manipulate these geometries in-place.

Once the geoms are readied, we can use tools such as nx_decompose for generating granular graph representations and nx_to_dual for casting a primal graph representation to its dual.

# this will (optionally) decompose the graph
G_decomp = graphs.nx_decompose(G, 50)
plot.plot_nx(G_decomp, plot_geoms=True, labels=False, dpi=200, figsize=(4, 4))

A decomposed graph.

# this will (optionally) cast to a dual network
G_dual = graphs.nx_to_dual(G)
# here we are plotting the newly decomposed graph (blue) against the original graph (red)
plot.plot_nx_primal_or_dual(G, G_dual, plot_geoms=False, dpi=200, figsize=(4, 4))

A dual graph (blue) plotted against the primal source graph (red). In this case, the true geometry has not been plotted so that the dual graph is easily delineated from the primal graph.

Metrics

After graph preparation and cleaning has been completed, the networkX graph can be transformed into data structures for efficiently computing centralities, land-use measures, or statistical aggregations.

Use network_structure_from_nx to convert a networkX graph into GeoPandas GeoDataFrames and a rustalgos.NetworkStructure, the latter of which is used by cityseer for efficiently computing the measures with the underlying rust algorithms.

Network Centralities

The networks.node_centrality_shortest, networks.node_centrality_simplest, and networks.segment_centrality methods wrap underlying rust functions that compute the centrality methods. All selected measures and distance thresholds are computed simultaneously to reduce the amount of time required for multi-variable and multi-scalar workflows. The results of the computations will be written to the GeoDataFrame.

from cityseer.metrics import networks

# create a Network layer from the networkX graph
# use a CRS EPSG code matching the projected coordinate reference system for your data
nodes_gdf, edges_gdf, network_structure = io.network_structure_from_nx(G_decomp, crs=3395)
# the underlying method allows the computation of various centralities simultaneously, e.g.
nodes_gdf = networks.segment_centrality(
    # the network structure for which to compute the measures
    network_structure=network_structure,
    # the nodes GeoDataFrame, to which the results will be written
    nodes_gdf=nodes_gdf,
    # the distance thresholds for which to compute centralities
    distances=[200, 400, 800, 1600],
)
# the results are now in the GeoDataFrame
nodes_gdf.head()

# plot centrality
from matplotlib import colors

# custom colourmap
cmap = colors.LinearSegmentedColormap.from_list("cityseer", ["#64c1ff", "#d32f2f"])
# normalise the values
segment_harmonic_vals = nodes_gdf["cc_metric_segment_harmonic_800"]
segment_harmonic_vals = colors.Normalize()(segment_harmonic_vals)
# cast against the colour map
segment_harmonic_cols = cmap(segment_harmonic_vals)
# plot segment_harmonic
# cityseer's plot methods are used here and in tests for convenience
# that said, rather use plotting methods directly from networkX or GeoPandas where possible
plot.plot_nx(G_decomp, labels=False, node_colour=segment_harmonic_cols, dpi=200, figsize=(4, 4))

800m segmentised harmonic centrality.

Land-use and statistical measures

Landuse and statistical measures require a GeoPandas GeoDataFrame consisting of Point geometries. Columns representing categorical landuse information ("pub", "shop", "school") can be passed to landuse methods, whereas columns representing numerical information can be used for statistical methods.

When computing these measures, cityseer will assign each data point to the two closest network nodes — one in either direction — based on the closest adjacent street edge. This enables cityseer to use dynamic spatial aggregation methods that more accurately describe distances from the perspective of pedestrians travelling over the network, and relative to the direction of approach.

layers.compute_landuses and layers.compute_mixed_uses methods are used for the calculation of land-use accessibility and mixed-use measures whereas layers.compute_stats can be used for statistical aggregations. As with the centrality methods, the measures are computed over the network and are computed simultaneously for all measures and distances.

from cityseer.metrics import layers

# a mock data dictionary representing categorical landuse data
# here randomly generated letters represent fictitious landuse categories
data_gdf = mock.mock_landuse_categorical_data(G_decomp, random_seed=25)
data_gdf.head()

# example easy-wrapper method for computing mixed-uses
# this is a distance weighted form of hill diversity
nodes_gdf, data_gdf = layers.compute_mixed_uses(
    # the source data
    data_gdf,
    # column in the dataframe which contains the landuse labels
    landuse_column_label="categorical_landuses",
    # nodes GeoDataFrame - the results are written here
    nodes_gdf=nodes_gdf,
    # measures will be computed relative to pedestrian distances over the network
    network_structure=network_structure,
    # distance thresholds for which you want to compute the measures
    distances=[200, 400, 800, 1600],
)
# the GeoDataFrame will contain the results of the calculations
print(nodes_gdf.columns)
# which can be retrieved as needed
print(nodes_gdf["cc_metric_q0_800_hill"])

# for curiosity's sake:
# plot the assignments to see which edges the data points were assigned to
plot.plot_assignment(network_structure, G_decomp, data_gdf, dpi=200, figsize=(4, 4))

Data points assigned to a Network Layer.

Data assignment becomes more precise on a decomposed Network Layer.

# plot distance-weighted "hill" numbers mixed uses
mixed_uses_vals = nodes_gdf["cc_metric_q0_800_hill_weighted"]
mixed_uses_vals = colors.Normalize()(mixed_uses_vals)
mixed_uses_cols = cmap(mixed_uses_vals)
plot.plot_assignment(
    network_structure,
    G_decomp,
    data_gdf,
    node_colour=mixed_uses_cols,
    data_labels=data_gdf["categorical_landuses"].values,
    dpi=200,
    figsize=(4, 4),
)

800m distance-weighted mixed-uses.

# compute landuse accessibilities for land-use types a, b, c
nodes_gdf, data_gdf = layers.compute_accessibilities(
    # the source data
    data_gdf,
    # column in the dataframe which contains the landuse labels
    landuse_column_label="categorical_landuses",
    # the landuse categories for which to compute accessibilities
    accessibility_keys=["a", "b", "c"],
    # nodes GeoDataFrame - the results are written here
    nodes_gdf=nodes_gdf,
    # measures will be computed relative to pedestrian distances over the network
    network_structure=network_structure,
    # distance thresholds for which you want to compute the measures
    distances=[200, 400, 800, 1600],
)
# accessibilities are computed in both weighted and unweighted forms
# e.g. for "a" and "b" landuse codes in weighted and non weighted, respectively
print(nodes_gdf[["cc_metric_a_800_weighted", "cc_metric_b_1600_non_weighted"]])

Aggregations can likewise be computed for numerical data. Let's generate some mock numerical data:

numerical_data_gdf = mock.mock_numerical_data(G_decomp, num_arrs=3)
numerical_data_gdf.head()
# compute stats for column mock_numerical_1
nodes_gdf, numerical_data_gdf = layers.compute_stats(
    # the source data
    numerical_data_gdf,
    # numerical column to compute stats for
    stats_column_label="mock_numerical_1",
    # nodes GeoDataFrame - the results are written here
    nodes_gdf=nodes_gdf,
    # measures will be computed relative to pedestrian distances over the network
    network_structure=network_structure,
    # distance thresholds for which you want to compute the measures
    distances=[800, 1600],
)
# statistical aggregations are calculated for each requested column,
# and in the following forms:
# max, min, sum, sum_weighted, mean, mean_weighted, variance, variance_weighted
print(nodes_gdf["cc_metric_max_800"])
print(nodes_gdf["cc_metric_mean_wt_800"])

The landuse metrics and statistical aggregations are computed over the street network relative to the network, with results written to each node. The mixed-use, accessibility, and statistical aggregations can therefore be compared directly to centrality computations from the same locations, and can be correlated or otherwise compared.

Data derived from metrics can be converted back into a NetworkX graph using the nx_from_cityseer_geopandas method.

nx_multigraph_round_trip = io.nx_from_cityseer_geopandas(
    nodes_gdf,
    edges_gdf,
)
nx_multigraph_round_trip.nodes["0"]