In teamookla/ooklaOpenDataR: Work with 'Speedtest by Ookla' Global Internet Performance Data

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>"
)

The global dataset is helpful, except when you aren't interested in dealing with global data. Often times, users have a specific area of interest in mind. The filter_by_quadkey() function is one way of dealing with this issue. The function takes a set of tiles (from get_performance_tiles()) and a bounding box (from st_bbox()) and returns just the tiles within the set bounding box.

There are two main ways of setting this box: a shapefile or manually input coordinates. We'll walk through both.

library(ooklaOpenDataR)
library(sf)
library(mapview) # check our tiles

Filter tiles by manual coordinates

Choosing coordinates from somewhere like Google maps is flexible and quick, but not necessarily reproducible or sustainable. When you don't have a boundary shapefile, though, it's a good alternative. We'll see how to do this with Google maps, but any source where you can find coordinates would work.

Let's say you're interested in the fixed broadband performance around Marrakesh, Morocco but don't have a shapefile of the boundary on hand. We can still approximate it!

Type in the place in the search bar: Marrakesh, Morocco

knitr::include_graphics("marrakesh_search.png")

Once you have the area highlighted, you can start choosing coordinates. We'll start at the top to get the ymax coordinate. You want to pick a point that's at least includes the northernmost edge of your area.

knitr::include_graphics("marrakesh_top.png")

I usually round my coordinates for simplicity so we'll say ymax = 31.72.

Next we'll repeat the process and click on the rightmost point to get the xmax coordinate.

knitr::include_graphics("marrakesh_right.png")

Based on this point we'll say xmax = -7.88. Repeat this process again for the bottom (ymin) and leftmost (xmin) coordinates.

Once you've gathered all the coordinates, you can create the bounding box with the following code:

bbox <- st_bbox(c(xmin = -8.13, xmax = -7.88, ymax =31.72, ymin = 31.55), crs = st_crs(4326)) # 4326 means that the coordinates are latitude and longitude

Pull the global dataset for fixed data in Q2 2020 and filter with the bbox object.

marrakesh_fixed_tiles <- get_performance_tiles(service = "fixed", year = 2020, quarter = 2, sf = TRUE) %>%
  filter_by_quadkey(bbox = bbox)

Then to check that the tiles are in the right place we'll use mapview() to make a quick interactive map with a basemap.

mapview(marrakesh_fixed_tiles)

Success!

Filter tiles by shapefile

If you do have a shapefile (or other geographic data file, geoJSON for example), the process is much simpler. We'll use Minneapolis, MN as an example here.

The city's open data portal has a boundary file we can use for filtering the tiles.

minneapolis <- read_sf("https://opendata.arcgis.com/datasets/89f1a70c0cf24d7692e2d02fdf8f4e47_0.geojson") %>%
  st_transform(4326) # transform to lat/lon

Then we'll download the fixed performance tiles for Q2 2020 and filter them with the Minneapolis boundary.

mpls_fixed_tiles <- get_performance_tiles(service = "fixed", year = 2020, quarter = 2, sf = TRUE) %>%
  filter_by_quadkey(bbox = st_bbox(minneapolis))

Check to make sure it looks right:

mapview(mpls_fixed_tiles)

Success!

teamookla/ooklaOpenDataR documentation built on Oct. 23, 2021, 9:48 a.m.