Getting started with ukpolice

Nicholas Tierney

2018-05-20

ukpolice is an R package that facilitates retrieving data from the UK police database.. The data provided by the API contains public sector information licensed under the Open Government Licence v3.0.

Get the crime data with ukp_crime

ukp_crime() draws crimes from within a one mile radius of the location.

library(ukpolice)

crime_data <- ukp_crime(lat = 52.629729, lng = -1.131592)
#> No encoding supplied: defaulting to UTF-8.

head(crime_data)
#> # A tibble: 6 x 12
#>   category  persistent_id date    lat  long street_id street_name  context
#>   <chr>     <chr>         <chr> <dbl> <dbl> <chr>     <chr>        <chr>  
#> 1 anti-soc… ""            2018…  52.6 -1.13 882346    On or near … ""     
#> 2 anti-soc… ""            2018…  52.6 -1.13 883351    On or near … ""     
#> 3 anti-soc… ""            2018…  52.6 -1.13 883408    On or near … ""     
#> 4 anti-soc… ""            2018…  52.6 -1.13 883313    On or near … ""     
#> 5 anti-soc… ""            2018…  52.6 -1.11 882420    On or near … ""     
#> 6 anti-soc… ""            2018…  52.6 -1.13 882336    On or near … ""     
#> # ... with 4 more variables: id <chr>, location_type <chr>,
#> #   location_subtype <chr>, outcome_status <chr>

When no date is specified, it uses the latest month available, which can be found using ukp_last_update().

ukp_last_update()
#> No encoding supplied: defaulting to UTF-8.
#> [1] "2018-03"

Specify a month and year for data

When date is specified, it must be in the format “YYYY-MM”. Currently ukp_crime() only allows for searching of that current month.

crime_data_date <- ukp_crime(lat = 52.629729, 
                             lng = -1.131592,
                             date = "2016-03")
#> No encoding supplied: defaulting to UTF-8.

head(crime_data_date)
#> # A tibble: 6 x 12
#>   category  persistent_id date    lat  long street_id street_name  context
#>   <chr>     <chr>         <chr> <dbl> <dbl> <chr>     <chr>        <chr>  
#> 1 anti-soc… ""            2016…  52.6 -1.14 882313    On or near … ""     
#> 2 anti-soc… ""            2016…  52.6 -1.12 883287    On or near … ""     
#> 3 anti-soc… ""            2016…  52.6 -1.15 883538    On or near … ""     
#> 4 anti-soc… ""            2016…  52.6 -1.13 883415    On or near … ""     
#> 5 anti-soc… ""            2016…  52.6 -1.14 883525    On or near … ""     
#> 6 anti-soc… ""            2016…  52.6 -1.14 883433    On or near … ""     
#> # ... with 4 more variables: id <chr>, location_type <chr>,
#> #   location_subtype <chr>, outcome_status <chr>

This is still a little buggy at the moment as it returns blank columns for variables like persistent_id and context, location_subtype, and outcome_status. This issue is currently logged at issue #11.

Example usage

Explore the number of crime types

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
library(ggplot2)

crime_data <- ukp_crime(lat = 52.629729, lng = -1.131592)
#> No encoding supplied: defaulting to UTF-8.

crime_data %>%
  count(category) %>%
  ggplot(aes(x = reorder(category, n),
             y = n)) + 
  geom_col() + 
  labs(x = "Crime Type",
       y = "Number of Crimes",
       title = paste0("Crimes commited in ",crime_data_date$date[1])) +
  coord_flip() +
  theme_minimal()

Use leaflet

You can add a popup that displays the crime type using the popup argument in leaflet.

library(leaflet)
crime_data <- ukp_crime(lat = 52.629729, lng = -1.131592)
#> No encoding supplied: defaulting to UTF-8.
crime_data %>%
  leaflet() %>%
  addTiles() %>%
  addCircleMarkers(popup = ~category)
#> Assuming "long" and "lat" are longitude and latitude, respectively

Get crime within a polygon

ukp_crime_poly() finds all crimes within the polygon provided by a dataframe with columns names “lat” and “long”.

poly_df_3 <- data.frame(lat = c(52.268, 52.794, 52.130),
                        long = c(0.543, 0.238, 0.478))

poly_df_3
#>      lat  long
#> 1 52.268 0.543
#> 2 52.794 0.238
#> 3 52.130 0.478

ukp_data_poly_3 <- ukp_crime_poly(poly_df_3)
#> No encoding supplied: defaulting to UTF-8.

head(ukp_data_poly_3)
#> # A tibble: 6 x 12
#>   category  persistent_id date    lat  long street_id street_name  context
#>   <chr>     <chr>         <chr> <dbl> <dbl> <chr>     <chr>        <chr>  
#> 1 anti-soc… ""            2018…  52.3 0.496 1141362   On or near … ""     
#> 2 anti-soc… ""            2018…  52.3 0.497 1141337   On or near … ""     
#> 3 anti-soc… ""            2018…  52.3 0.476 1140354   On or near … ""     
#> 4 anti-soc… ""            2018…  52.3 0.413 557698    On or near … ""     
#> 5 anti-soc… ""            2018…  52.3 0.413 564411    On or near … ""     
#> 6 anti-soc… ""            2018…  52.2 0.484 561970    On or near … ""     
#> # ... with 4 more variables: id <chr>, location_type <chr>,
#> #   location_subtype <chr>, outcome_status <chr>

Neighbourhood information

ukp_neighbourhood(), retrieves a list of neighbourhoods for a force, https://data.police.uk/docs/method/neighbourhoods/

This returns a tibble with columns id and name.

ukp_neighbourhood("leicestershire")
#> No encoding supplied: defaulting to UTF-8.
#> # A tibble: 67 x 2
#>    id    name                          
#>    <chr> <chr>                         
#>  1 NC04  City Centre                   
#>  2 NC66  Cultural Quarter              
#>  3 NC67  Riverside                     
#>  4 NC68  Clarendon Park                
#>  5 NE09  Belgrave South                
#>  6 NE10  Belgrave North                
#>  7 NE11  Rushey Mead                   
#>  8 NE12  Humberstone                   
#>  9 NE13  Northfields, Tailby and Morton
#> 10 NE14  Thurncourt                    
#> # ... with 57 more rows

• id is a Police force specific team identifier, (note that this identifier is not unique and may also be used by a different force).
• name is the name for the neighbourhood.