Data Packs Documentation

This vignette acts as extended documentation of our transport infrastructure data packs for the following content:

The contents of this vignette are as follows: * Workflow for data pack production * Example data pack production (Leeds) * Openinfra function definitions and descriptions

Data Pack Production Workflow

Openinfra tranport infrastructure data packs are produced using the following, reproducible, workflow:

1. Download OSM data file for entirety of England
2. For each LAD polygon, get the OSM points & lines data from the previously downloaded England file & save each LAD default points lines network locally
3. For each LAD points & lines network saved locally, load the network and apply the openinfra functions, recategorising the OSM data
4. Save the newly created points & lines data packs locally, before uploading to GitHub releases
5. Repeat 2 - 4 for each LAD within England

The workflow described above is contained within the create_data_packs.R script.

For ease of use, and segmenting long computation times, the above script has been split in two - create_LAD_networks.R performing steps 1-2, and create_upload_data_packs.R for steps 3-4.

Example Data Pack Production - Leeds

Here we cover a reproducible example of the creation of a transport infrastructure data pack from OSM data retrieved using osmextract, the outputs of which are visualised above.

This example concerns the LAD of Leeds, UK.
Note that a 2km circular buffer has been applied (centered at Long, Lat coords (-1.548567, 53.801277)) to reduce filesizes of the visualisations above.

Getting Data

# Data downloaded through osmextract as shown below: 

#total_place = osmextract::oe_get(
#  place = "Leeds",
#  provider = "bbbike",
#  layer = "lines",
#  never_skip_vectortranslate = TRUE,
#  force_download = TRUE,
#  quiet = FALSE,
#  extra_tags = c("foot", "bicycle", "access", "service", "maxspeed", "oneway",
#                  "kerb", "footway", "sidewalk", "cycleway", "segregated", 
#                  "highway", "crossing", "lit", "tactile_paving", "surface", 
#                  "smoothness", "width", "est_width", "lit_by_led", "ref", 
#                  "amenity")
#)

# For reproduciblity you may chose to use the example dataset from our releases.
# You can create your own data using the oe_get() function above.
leeds_lines_network = sf::read_sf("https://github.com/udsleeds/openinfra/releases/download/0.4.2/leeds_lines_network.geojson")
leeds_points_network = sf::read_sf("https://github.com/udsleeds/openinfra/releases/download/0.4.2/leeds_points_network.geojson")

Applying Openinfra Functions - Adding Value

# First we apply relevant functions to the "lines" (roads, paths, etc.) layer 
leeds_lines_datapack = oi_active_cycle(leeds_lines_network)
leeds_lines_datapack = oi_active_walk(leeds_lines_datapack)
leeds_lines_datapack = oi_clean_maxspeed_uk(leeds_lines_datapack)
leeds_lines_datapack = oi_cycle_crossings(leeds_lines_datapack)
leeds_lines_datapack = oi_cycle_separation(leeds_lines_datapack)
leeds_lines_datapack = oi_is_lit(leeds_lines_datapack)
leeds_lines_datapack = oi_recode_road_class(leeds_lines_datapack)
leeds_lines_datapack = oi_road_names(leeds_lines_datapack)

# Now we create points data packs
leeds_points_datapack = oi_bicycle_parking(leeds_points_network)

Selecting Relevant Columns for Data Pack

# Select relevant lines columns
leeds_lines_datapack = leeds_lines_datapack %>% dplyr::select(c(
 "osm_id", "highway", matches(match = "openinfra_|im_")))

# Append geometry as final column - good {sf} practice.
leeds_lines_datapack = sf::st_sf(
  leeds_lines_datapack %>% sf::st_drop_geometry(),
  geometry = leeds_lines_datapack$geometry)

# Select relevant points columns
leeds_points_datapack = leeds_points_datapack %>% dplyr::select(c(
  "osm_id", matches(match="openinfra_|im_")))

# Append geometry as final column - good {sf} practice.
leeds_points_datapack = sf::st_sf(
  leeds_points_datapack %>% sf::st_drop_geometry(),
  geometry = leeds_points_datapack$geometry)

Example visualisation

library(dplyr)

## 
## Attaching package: 'dplyr'

## The following objects are masked from 'package:stats':
## 
##     filter, lag

## The following objects are masked from 'package:base':
## 
##     intersect, setdiff, setequal, union

# Get data for cell
leeds_points_datapack = sf::read_sf("https://github.com/udsleeds/openinfra/releases/download/0.4.2/leeds_points_datapack.geojson")
# Remove NA 
leeds_points_datapack = leeds_points_datapack %>% dplyr::filter(! is.na(openinfra_cycle_parking)) 

#Visualise Data
tmap::tmap_mode("view")

## tmap mode set to interactive viewing

tmap::tm_shape(leeds_points_datapack) + 
  tmap::tm_bubbles(col = "openinfra_cycle_parking", size=0.1)

This concludes the reproducible example creation of the transport infrastructure data packs.

Below we see how each of the openinfra functions has been defined, thus the processing steps applied to OSM data in data pack creation.

Data Packs: openinfra Function Definitions

The openinfra package functions used to create the transport infrastructure data packs have documentation available here. The source code for each function can be found in project GitHub R/ directory.

Data Packs Documentation

James Hulse

2022-12-14