Introduction to R for Road Safety

In stats19: Work with Open Road Traffic Casualty Data from Great Britain

background-image: url(https://user-images.githubusercontent.com/1825120/61066243-12dc6d80-a3fd-11e9-8805-826a47c553f6.png)

About the course team and location

• Robin Lovelace
• Haruko
• Pangiotis
• Martyna

---

About the course

• What it is...
• An opportunity to learn R with support
• Based on recent, cutting edge software
• By people experienced teachers/researchers/developers

--

• What it is not
• A course on statistical modelling

---

About the package

See https://docs.ropensci.org/stats19/

---

R and RStudio demo

Actions speak louder than words...

---

Over to you

Work through Section 2 of the exercises

• Projects and scripts
• Writing and running code
• Vewing Objects
• Autocompletion
• Getting help

---

Time data

Section 3.4 shows how to change classes in R

Time is represented by special classes:

library(lubridate)
x = today()
class(x)
day(x)
month(x)
year(x)
weekdays(x)

---

Time representations

as.Date("2019-10-17") # works
# as.Date("2019 10 17") # fails
ymd("2019 10 17") # works
dmy("17/10/2019") # works

---

Subsetting time objects

c_sample = stats19::accidents_sample
c_sample$date
c_sample$date_formatted = dmy(c_sample$date)
c_sample$date_formatted > ymd("2022-08-01")

---

Next up

• Try running the code and answer the questions in Section 5
• Continue working through Sections 2 and 3

After lunch

Spatial data (maps!)

Spatial data and maps

library(sf)
crashes = readr::read_csv("https://github.com/ropensci/stats19/releases/download/1.0.0/crashes.csv")
crashes_sf = crashes # create copy of crashes dataset
crashes_sf$longitude = c(-1.3, -1.2, -1.1)
crashes_sf$latitude = c(50.7, 50.7, 50.68)
crashes_sf = st_as_sf(crashes_sf, coords = c("longitude", "latitude"), crs = 4326)
plot(crashes_sf)

---

Doing spatial data with R

• Context: spatial ecosystem (see section 1.4 of Geocomputation with R - package ecosystem)

• Exercises: Section 6 of the handout

• Further reading: Section 3.2 to 3.2.2 of handouts

---

Practical

• Work through Section 6.1 and try exercises 1 to 3
• Take a short read of Section 1.4 of Geocomputation with R (first challenge, find the resource online!)
• Return to the practical

---

Afternoon session

• Talk on Road Safety 1

• Applying the methods to stats19 data - live demo

  How to access data with stats19 Key stats19 functions Excercises: analysing road crash data on the Isle of Wight

• Continue working through the handouts

• Talk on Road Safety 2
• Homework - ensure that you have at least read-over the handout Sections 1:5

---

Day 2

• Prioirty: consolidate knowledge from day 1
• Continuing to work through the handouts starting on Section 6
• Demonstrations of counting number of crashes on roads
• Talk on network analysis
• Talk on R in a professional setting
• Working on your own data

--

• Missed homework! Ensure that you have at least read-over the handout Sections 1:5

---

The R learning curve

See video: https://www.youtube.com/watch?v=7oyiPBjLAWY&feature=youtu.be&t=357

---

Agenda

09:30-11:00 Point pattern analysis

• Visualising data with tmap
• Spatial and temporal subsetting
• Aggregation

11:15-12:30 Road network data

• Desire lines: using origin-destination data
• Downloading road network data from OSM
• Buffers on road networks

Lunch

13:30-15:00 Analysing crash data on road network

Break

15:15-15:30: Talk on Road Safety 3

15:30-16:30 Applying the methods to your own data

---

Bonus extras

--

Merging, forecasting, network analysis

---

Visualising spatial data practical

• Foundations of sf: try Sections 6.2 and 6.3 (building on new subsetting knowledge)

• Making maps: try out the exercises in Section 7

• From break until lunch, either:

• Continue working through the practicals (Sections 2 to 5), or

• Dive into processing large stats19 dataset (Section 8)

• Sections 1:5 R vital

• Section 3 on subsetting (vital)
• Section 4 on packages (important)
• Section 5 on times (can do later)

--

• Live demo + questions

---

Further information

stats19 documentation built on Nov. 5, 2023, 1:09 a.m.