Package 'ttcalib'

Calibrate travel times to network centrality.

Description

This calibration step is performed after calibration to waiting-time penalties for traffic lights and turning across oncoming traffic, performed with the ttcalib_penalties function. For Santiago, that function gives an optimal waiting time at traffic lights of 16 seconds, and waiting time to turn across oncoming traffic of 1 second. These values should be used to generate a weighted network with time-based centrality via the ttcalib_streetnet function.

Usage

ttcalib_centrality(
  path_graph,
  path_uberdata,
  city = "santiago",
  hours = c(7, 10),
  turn_penalty = 1
)

Arguments

path_graph	Path to locally-saved fst-format weighted street network including centrality column.
path_uberdata	Path to Uber movement data.
city	Currently only accepts "santiago"
hours	A vector of two values defining the range of hours for Uber Movement data to be filtered. Value of NULL aggregates all hours without filtering.
turn_penalty	The value of the time penalty for waiting to turn across oncoming traffic used to generate the graph stored at path.

Details

The result of that call is presumed to have been saved using the fst package (with write_fst), which strips all attributes of the graph. These attributes must then be manually re-instated, and so are required to be submitted as parameters to this function.

---

Read geometries for chosen city from Uber Movement data.

Description

Currently hard-coded to Brussels or Santiago data only.

Usage

ttcalib_geodata(path, city = "santiago")

Arguments

path	Path to directory containing Uber movement data.
city	One of "brussels" or "santiago" (case-insensitive).

Value

A 'data.frame' with centroids of polygons used to aggregate movement data.

---

Calculate standard error against 'uber' data for all weighted networks generated via ttcalib_streetnet_batch.

Description

Calculate standard error against 'uber' data for all weighted networks generated via ttcalib_streetnet_batch.

Usage

ttcalib_penalties(path_results, path_uberdata, city, hours = NULL)

Arguments

path_results	Path to directory holding main OSM network data, including a sub-directory with output of ttcalib_streetnet_batch.
path_uberdata	Path to Uber movement data.
city	One of "brussels" or "santiago" (case-insensitive).
hours	A vector of two values defining the range of hours for data to be filtered. Default of NULL returns aggregate of all hours without filtering.

---

Load an 'SC' street network, weight for motorcar routing, and optionally calculate centrality.

Description

Load an 'SC' street network, weight for motorcar routing, and optionally calculate centrality.

Usage

ttcalib_streetnet(
  path,
  centrality = FALSE,
  penalty_traffic_lights = 8,
  penalty_turn = 7.5,
  dist_threshold = 10000
)

Arguments

path	Path to 'SC'-format file containing street network data.
centrality	If TRUE, calculate network centrality on all graph edges. Load an 'SC' street network, weight for motorcar routing, and calculate centrality.
penalty_traffic_lights	Time penalty for waiting at traffic lights (in seconds).
penalty_turn	Time penalty for turning across oncoming traffic.
dist_threshold	Threshold used for centrality calculations (in metres); see documentation for dodgr function, 'dodgr_centrality' for information.

Value

Network with centrality estimates on each edge.

---

Weight an 'SC' street network by a range of traffic light penalties, and locally save the results.

Description

This function produces an array of differently-weighted streetnets, each locally saved to a uniquely-named file. The resultant graphs are saved with the fst package, which removes the necessary attributes of the resultant data.frame objects. These must be manually restored prior to submitting to ttcalib_traveltimes. Required attributes are:

• "left_side", passed as same parameter to dodgr function wt_streetnet.

• "wt_profile", passed as same parameter to dodgr function wt_streetnet, and which should generally be "motorcar" here.

Usage

ttcalib_streetnet_batch(path, penalty_traffic_lights = 1:10)

Arguments

path	Path to 'SC'-format file containing street network data.
penalty_traffic_lights	Time penalty for waiting at traffic lights (in seconds).

Note

This function only generated weighted networks for a range of traffic light penalties. Turn penalties are calculated on-the-fly in m4ra for each routing call, and can be specified by modifying the "turn_penalty" attribute of the graph.

---

Estimate travel times between all Uber movement polygons

Description

Estimate travel times between all Uber movement polygons

Usage

ttcalib_traveltimes(graph, geodata, uberdata)

Arguments

graph	A dodgr graph returned from ttcalib_streetnet.
geodata	A data.frame returned from ttcalib_geodata.
uberdata	A data.frame returned from ttcalib_uberdata.

Value

A 'data.frame' with columns of m4ra estimates of travel times and corresponding empirical values from Uber movement data.

---

Read Uber Movement data and filter to defined time limits.

Description

Currently hard-coded to Brussels data only.

Usage

ttcalib_uberdata(path, city = "santiago", hours = NULL)

Arguments

path	Path to directory containing Uber movement data.
city	One of "brussels" or "santiago" (case-insensitive).
hours	A vector of two values defining the range of hours for data to be filtered. Default of NULL returns aggregate of all hours without filtering.

Value

A 'data.frame' of Uber Movement estimates of travel times.

---