In nacnudus/nzpullover: Driving Offences in New Zealand Between 2009 and 2017

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  fig.path = "figures/README-"
)

nzpullover

The nzpullover package makes available road policing data from the New Zealand Police in a convenient 'tidy' form. Offence categories include speed, alcohol, red light, restraints (seatbelts), mobile phone, fleeing drivers, and speeding police vehicles.

The original data published by the New Zealand Police is available as a spreadsheet here: http://www.police.govt.nz/about-us/publication/road-policing-driver-offence-data-january-2009-june-2018. Please note that this R package is not associated with the New Zealand Police.

The spreadsheet data has been made tidy by using the tidyxl and unpivotr packages. See the scripts in /data-raw. Zipped csv files are available in /inst/extdata.

The New Zealand Police refresh the data approximately quarterly. If this package lags behind, then please open an issue https://github.com/nacnudus/nzpullover/issues. The CRAN version will be updated annually.

Installation

You can install nzpullover from github with:

# install.packages("devtools")
devtools::install_github("nacnudus/nzpullover")

Example

Scroll to the bottom for lots more graphs.

library(nzpullover)
library(tidyverse)
library(forcats)

plotdata <-
  excess %>%
  mutate(series = fct_recode(series,
                             `Officer-issued` = "Officer-issued excess speed band",
                             `Camera-issued` =  "Camera-issued excess speed band")) %>%
  filter(series %in% c("Officer-issued",
                       "Camera-issued")) %>%
  group_by(series, month, speed) %>%
  summarise(value = sum(value)) %>%
  group_by(series, month) %>%
  mutate(prop = value / sum(value),
         cumprop = cumsum(prop))

plotdata %>%
  ggplot(aes(speed, value, colour = month)) +
  geom_point(shape = "-", size = 20) +
  scale_colour_continuous(name = "", trans = "date") +
  scale_y_continuous(labels = scales::comma) +
  facet_wrap(~series, nrow = 1) +
  ggtitle("New Zealand summer speed limit infringements/offences",
          subtitle = "Frequency in each speed band") +
  xlab("Speed band (km/h over the limit)") +
  ylab("Frequency")

plotdata %>%
  ggplot(aes(speed, cumprop, colour = month)) +
  geom_point(shape = "-", size = 20) +
  scale_colour_continuous(name = "", trans = "date") +
  scale_y_continuous(labels = scales::percent) +
  facet_wrap(~series, nrow = 1) +
  ggtitle("New Zealand summer speed limit infringements/offences",
          subtitle = "Cumulative proportion in each speed band") +
  xlab("Speed band (km/h over the limit)") +
  ylab("Cumulative proportion")

glimpse(driving_offences)
glimpse(static_camera)
glimpse(excess)
glimpse(fleeing_area)
glimpse(fleeing_district)
glimpse(police_speeding)
glimpse(police_speeding_band)

Other New Zealand datasets by the same author:

nzlifetables -- cohort life tables
nzcensus2013 -- summary statistics from the 2013 New Zealand census
nzbabynames -- the top 100 baby names annually since 1954, and annual birth statistics.

Lots more graphs

library(tidyverse)
library(fs)
library(nzpullover)

theme_set(theme_grey() + theme_minimal())

all_months <- tibble(month = seq(min(driving_offences$month),
                                 max(driving_offences$month),
                                 by = "month"))

timeseries <- function(series) {
  series_plot <-
  driving_offences %>%
    filter(series == rlang::UQ(series)) %>%
    group_by(month) %>%
    summarise(value = sum(value)) %>%
    right_join(all_months, by = "month") %>%
    ggplot(aes(month, value)) +
    geom_line() +
    geom_point() +
    scale_y_continuous(limits = c(0, NA), labels = scales::comma) +
    xlab("") +
    ylab("") +
    ggtitle(str_wrap(series))
  print(series_plot)
}

driving_offences %>%
  distinct(series) %>%
  pull(series) %>%
  walk(timeseries)