tidytransit (version 1.7.1)

cluster_stops: Cluster nearby stops within a group

Description

Finds clusters of stops for each unique value in group_col (e.g. stop_name). Can be used to find different groups of stops that share the same name but are located more than max_dist apart. gtfs_stops is assigned a new column (named cluster_colname) which contains the group_col value and the cluster number.

Usage

cluster_stops(
  gtfs_stops,
  max_dist = 300,
  group_col = "stop_name",
  cluster_colname = "stop_name_cluster"
)

Value

Returns a stops table with an added cluster column. If gtfs_stops is a tidygtfs object, a modified tidygtfs object is return

Arguments

gtfs_stops

Stops table of a gtfs object. It is also possible to pass a tidygtfs object to enable piping.

max_dist

Only stop groups that have a maximum distance among them above this threshold (in meters) are clustered.

group_col

Clusters for are calculated for each set of stops with the same value in this column (default: stop_name)

cluster_colname

Name of the new column name. Can be the same as group_col to overwrite.

Details

stats::kmeans() is used for clustering.

Examples

# \donttest{
library(dplyr)
nyc_path <- system.file("extdata", "nyc_subway.zip", package = "tidytransit")
nyc <- read_gtfs(nyc_path)
nyc <- cluster_stops(nyc)

# There are 6 stops with the name "86 St" that are far apart
stops_86_St = nyc$stops %>% 
  filter(stop_name == "86 St")

table(stops_86_St$stop_name_cluster)

stops_86_St %>% select(stop_id, stop_name, parent_station, stop_name_cluster) %>% head()

library(ggplot2)
ggplot(stops_86_St) +
  geom_point(aes(stop_lon, stop_lat, color = stop_name_cluster))
# }