AnalyseNetwork

analyse_network.AnalyseNetwork(self, gdf, osm, gtfs, out_path)

Class to calculate transport network using r5py.

Parameters

Name	Type	Description	Default
`gdf`	gpd.GeoDataFrame	Geodataframe with the population centroids.	required
`osm`	Union[str, pathlib.Path]	Path to the location of the open street map file.	required
`gtfs`	list	List including path or paths to the locations of gtfs files.	required
`out_path`	Union[str, pathlib.Path]	Path to save the output as parquet files.	required

Attributes

Name	Type	Description
gdf	gpd.GeoDataFrame	Geodataframe with the population centroids.
out_path	Union[str, pathlib.Path]	Path to save the output as parquet files.
transport_network	TransportNetwork	R5py object that contains a transport network initialised with data from OpenStreetMap and GTFS.

Raises

Type Description

NotImplementedError If the out_path provided already contains parquet files, there is a risk that if the number of parquet files produced is less than the number of files in the directory not all of them will be overwritten. This will cause issues when loading them as there will be additional records in the dataframe. In the future we will implement something to handle this (emptying folder, archiving old files). In the meantime, this will raise an error.

Type	Description
NotImplementedError	If the `out_path` provided already contains parquet files, there is a risk that if the number of parquet files produced is less than the number of files in the directory not all of them will be overwritten. This will cause issues when loading them as there will be additional records in the dataframe. In the future we will implement something to handle this (emptying folder, archiving old files). In the meantime, this will raise an error.

Methods

Name	Description
od_matrix	Calculate full O-D matrix and save as parquet.

od_matrix

analyse_network.AnalyseNetwork.od_matrix(batch_orig=False, partition_size=200, destination_col='within_urban_centre', distance=11.25, unit=Unit.KILOMETERS, **kwargs)

Calculate full O-D matrix and save as parquet.

Parameters

Name	Type	Description	Default
`batch_orig`	bool	Flag to indicate whether to calculate the transport network performance using the whole dataset or batching by origin (iteratively calculating the transport performance of each origin and all destinations within range). Should be True for large urban centres where the whole Origin-Destination (O-D) matrix may be too large to be held in memory. Defaults to False.	`False`
`partition_size`	int	Maximum size of each individual parquet file. If data exceeds this size, it will be split in several parquet files.	`200`
`destination_col`	str	Column indicating what centroids should be considered as destinations. Default is “within_urban_centre”.	`'within_urban_centre'`
`distance`	Union[int, float]	Distance to filter destinations in km. Points further away from origin are removed from output. Default is 11.25 km.	`11.25`
`unit`	Unit	Unit to calculate distance. Default is km.	`Unit.KILOMETERS`
`**kwargs`		Any valid argument for r5py TravelTimeMatrixComputer object. E.g. `departure`, `max_time`, `transport_modes`.	`{}`

Notes

batch_orig = True can be used for small datasets, however this will likely be slower as normal r5py behaviour uses parallel computing. We recommend setting this argument to True only in cases where memory may not be enough to handle the O-D matrix.