This is to prepare the network for the search of shortest paths between node pairs. The computed weights can account for edge lenghts, distance from a target geometry, and whether or not an edge falls within a specified region, which we aim to exclude from search of the shortest paths.
Usage
add_weights(
network,
target = NULL,
exclude_area = NULL,
penalty = 1000,
weight_name = "weight"
)Arguments
- network
A network object
- target
Target geometry to calculate distances from, as a simple feature geometry
- exclude_area
Area that we aim to exclude from the shortest-path search, as a simple feature geometry
- penalty
Penalty (in the network CRS' units) that is added to the edges that falls within the excluded area
- weight_name
Name of the column in the edge table where to add the weights
Details
For the i-th edge of the network, its weight \(w_i\) is defined in the
following way:
$$
w_i = |e_i| + d_{geom}(e_i) + p_{buf}(e_i)
$$
where the first term is the edge length, the second one is the distance
from a target geometry (target, optional) and the last one is a penalty
that is added if the centroid of the edge falls within a specified region
(exclude_area, optional).
Shortest paths calculated on the resulting network will thus tend to prefer
edges close to target and to avoid edges within exclude_area.