Skip to content

Latest commit

 

History

History
269 lines (224 loc) · 10.9 KB

README.md

File metadata and controls

269 lines (224 loc) · 10.9 KB
Raw

AccessUK

logo

Lifecycle: experimental

AccessUK is an open-source tool for the R programming language designed to streamline the integration of accessibility measures into R workflows focusing on Great Britain (GB), thereby unlocking the potential of spatial data by informing decision-making processes and supporting sustainable practices across diverse sectors.

The main aim of this tool is to manage and distribute a series of pre-computed Accessibility-related measures for small geographic areas in GB described here (PTAI dataset). The pre-computed measures include accessibility to employmnet, heath, education, food services, and urban centres. It offers additional functions that enable users to customise or create new accessibility measures.

AccessUK uses DuckDB, a relational database management system optimised for analytical workloads, in the background. This facilitates the rapid execution of queries without loading the full data into virtual memory. This mitigates common computational challenges associated with large travel matrices.

Installation

To install the latest version of AccessUK directly from GitHub, please follow the steps below:

1. Install the devtools package (if not already installed)

You’ll need the devtools package to install R packages from GitHub. If you don’t have it installed, run:

install.packages("devtools")

2. Install AccessUK

With devtools installed, you can now install the AccessUK package:

devtools::install_github("urbanbigdatacentre/AccessUK")

Usage

AccessUK does three main things: (1) get ready-to-use accessibility measures for a range of urban and regional services; (2) tailor accessibility measures to destinations not previously included in the dataset based on pre-computed travel matrices; (3) estimate new accessibility measures based on user-generated TTMs and service locations.

library(AccessUK)
library(tidyverse)
library(sf)

1. Getting precomputed accessibility measures

To get precomputed measures, the first step is to define a list 2011 lower super output areas (LSOA) or data zones (DZ) code for the desired locations. the example below selects a few areas in central London,

lsoa_london <- c('E01000919', 'E01002726', 'E01003014', 'E01033490', 'E01000914', 'E01000937', 'E01032583', 'E01000855', 'E01003016', 'E01000918', 'E01000002', 'E01032739', 'E01000916', 'E01004734', 'E01000936', 'E01004735', 'E01002724', 'E01000850', 'E01033596', 'E01000852', 'E01000853', 'E01003927', 'E01033595', 'E01004731', 'E01000001', 'E01003935', 'E01004763', 'E01004736', 'E01000917', 'E01003929', 'E01003013', 'E01032740', 'E01032582', 'E01000915', 'E01003930', 'E01000920', 'E01003017', 'E01003934', 'E01004733', 'E01032584')

The desired locations are passed to the get_accessibility() function. Also, the type of destination can be specified in the ‘service’ argument. It can be chosen from “employment”, “supermarkets”, “school”, “gp” (general practice), “hospital”, and “cities”. The example below retrieves accessibility to employment in central London.

accessibility_london <- get_accessibility(
  from = lsoa_london, 
  service = 'employment', 
  mode = "public_transport"
)
glimpse(accessibility_london)
## Rows: 40
## Columns: 18
## $ geo_code           <chr> "E01032584", "E01032583", "E01032582", "E01003927",…
## $ geo_label          <chr> "Southwark 034E", "Southwark 034D", "Lambeth 036E",…
## $ employment_15      <dbl> 116600, 146950, 106650, 140075, 475675, 326450, 201…
## $ employment_30      <dbl> 2006530, 2265690, 2120155, 1924470, 1930935, 216792…
## $ employment_45      <dbl> 3382235, 3474700, 3508980, 3219590, 3187045, 325164…
## $ employment_60      <dbl> 4532830, 4630565, 4701855, 4311400, 4307635, 452590…
## $ employment_75      <dbl> 5937880, 5992060, 6095340, 5729995, 5717505, 599671…
## $ employment_90      <dbl> 7356745, 7474240, 7465010, 7140865, 7082720, 756973…
## $ employment_105     <dbl> 8488575, 8515250, 8598215, 8312200, 8297765, 879248…
## $ employment_120     <dbl> 9596665, 9628955, 9690610, 9379390, 9323300, 992857…
## $ employment_pct_15  <dbl> 0.3877880, 0.4887260, 0.3546963, 0.4658611, 1.58199…
## $ employment_pct_30  <dbl> 6.673313, 7.535226, 7.051206, 6.400398, 6.421899, 7…
## $ employment_pct_45  <dbl> 11.24863, 11.55615, 11.67016, 10.70770, 10.59947, 1…
## $ employment_pct_60  <dbl> 15.07528, 15.40032, 15.63742, 14.33884, 14.32632, 1…
## $ employment_pct_75  <dbl> 19.74819, 19.92838, 20.27187, 19.05680, 19.01526, 1…
## $ employment_pct_90  <dbl> 24.46705, 24.85781, 24.82711, 23.74907, 23.55569, 2…
## $ employment_pct_105 <dbl> 28.23128, 28.32000, 28.59592, 27.64470, 27.59669, 2…
## $ employment_pct_120 <dbl> 31.91657, 32.02396, 32.22901, 31.19395, 31.00741, 3…

The results include both the absolute employment available for a range of travel times in minutes as well as the relative employment in percent.

2. Tailored accessibility measures

Given new point locations in GB, the user can tailor its own accessibility measures at the LSOA/DZ level. This uses 2011 LSOA/DZ geometries in the background. This is combined with pre-computed travel times in the PTAI dataset. The example below uses retail points and for various time cuts.

# New destination points for this example
data_dir = system.file('data', package = "AccessUK")
load(file.path(data_dir, 'retail_points.RData'))
retail_points <- retail_points %>% 
  st_as_sf(coords = c('long_wgs', 'lat_wgs'), crs = 4326)

# Run my_accessibility
timecuts <- c(10, 20, 30)
retail_accessibility <- my_accessibility(
  destinations = retail_points, 
  time_cut = timecuts
)
## Warning: attribute variables are assumed to be spatially constant throughout
## all geometries

glimpse(retail_accessibility)
## Rows: 41,729
## Columns: 4
## $ from_id     <chr> "E01000001", "E01000002", "E01000003", "E01000005", "E0100…
## $ access_n_10 <dbl> 6, 4, 6, 0, 0, 3, 0, 3, 4, 0, 0, 2, 2, 2, 0, 2, 2, 0, 0, 1…
## $ access_n_20 <dbl> 27, 16, 25, 24, 5, 11, 6, 11, 11, 9, 8, 4, 3, 2, 3, 6, 3, …
## $ access_n_30 <dbl> 109, 74, 89, 95, 12, 15, 14, 17, 17, 13, 12, 18, 13, 9, 8,…

The result includes accessibility for all of GB, given that the retail points cover all the territory. However, we visualise a subset in central London.

# Read LSOA geometries
lsoa_geoms <- st_read(file.path(data_dir, 'lsoa_geoms/infuse_lsoa_lyr_2011_clipped.shp'))
## Reading layer `infuse_lsoa_lyr_2011_clipped' from data source 
##   `C:\Users\jvt3d\AppData\Local\Programs\R\R-4.3.0\library\AccessUK\data\lsoa_geoms\infuse_lsoa_lyr_2011_clipped.shp' 
##   using driver `ESRI Shapefile'
## Simple feature collection with 42619 features and 3 fields
## Geometry type: MULTIPOLYGON
## Dimension:     XY
## Bounding box:  xmin: -69.1254 ymin: 5337.9 xmax: 655604.7 ymax: 1220302
## Projected CRS: OSGB36 / British National Grid
# Map
retail_accessibility %>% 
  filter(from_id %in% lsoa_london) %>% 
  left_join(lsoa_geoms, by = c('from_id' = 'geo_code')) %>% 
  st_as_sf() %>% 
  ggplot() +
  geom_sf(aes(fill =  access_n_30), col = NA) +
  scale_fill_viridis_c() +
  labs(
    title = 'Accessibility to retail points in London by public transport within 30 minutes',
    subtitle = 'Contains retail points larger than 280 m2',
    fill = 'Retail points'
  ) +
  theme_void()

3. Compute new accessibility measures

The user can estimate completely new accessibility measures using DuckDB in the background for any region by providing a travel matrix in CSV or Parquet format. The user should input aggregated locations using the same ID as the travel matrix. This can contain different types of locations in each column with the first column named “id”.

# Aggregate locations in Manchester by LSOA
aggregated_retail <- retail_points %>% 
  filter(town == 'Manchester') %>% 
  my_accessibility(0)
## Warning: attribute variables are assumed to be spatially constant throughout
## all geometries
# First column as 'id'
aggregated_retail <- aggregated_retail %>% 
  rename(retail = access_n_0, id = from_id)

# Specify the location of a travel matrix. This can be a folder with multiple files or a single file
ttm_path <- file.path(data_dir, 'accessibility_indicators_gb/ttm/ttm_pt_20211122.csv')

# Estimate accessibility for various thresholds
new_accessibility <- estimate_accessibility(
  travel_matrix = ttm_path, 
  travel_cost =  'travel_time_p50',
  weights = aggregated_retail, 
  time_cut = timecuts
)
glimpse(new_accessibility)
## Rows: 41,729
## Columns: 4
## $ from_id          <chr> "E01000001", "E01000002", "E01000003", "E01000005", "…
## $ access_retail_10 <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ access_retail_20 <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ access_retail_30 <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,…

This returns all LSOA/DZ in GB. As we are providing data for Manchester only, we can filter relevant observations only.

new_accessibility %>% 
  filter(access_retail_20 > 0) %>% 
  left_join(lsoa_geoms, by = c('from_id' = 'geo_code')) %>% 
  st_as_sf() %>% 
  ggplot() +
  geom_sf(aes(fill =  access_retail_20), col = NA) +
  scale_fill_viridis_c() +
  labs(
    title = 'Accessibility to retail points in Manchester by public transport within 20 minutes',
    subtitle = 'Contains retail points larger than 280 m2',
    fill = 'Retail points'
  ) +
  theme_void()

Citation

If you use this package, please cite as following:

Related resources and projects

References

  1. Verduzco Torres, J. R., & McArthur, D. (2022). Public Transport Accessibility Indicators for Great Britain [dataset]. Zenodo. https://doi.org/10.5281/zenodo.8037156
  2. Verduzco Torres, J. R., & McArthur, D. P. (2024). Public transport accessibility indicators to urban and regional services in Great Britain. Scientific Data, 11(1), Article 1. https://doi.org/10.1038/s41597-023-02890-w