In wsteenbeek/NearRepeat: Near Repeat Calculation using the Knox Test (Monte Carlo permutation)

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

In order to illustrate the application of the tools contained here, an example dataset is provided. This vignette shows some pre-processing steps to derive a dataset in suitable form for analysis.

The data will be drawn from the Crime Open Database (CODE), maintained by Matt Ashby.[^1] This collates crime data from a number of open sources in a harmonized format. The spatial data is provided in lat/lon format; here the sp and rgdal libraries will be used to re-project the coordinates to metric units for distance calculations.

# This block is not run as it takes a while to download, but code should
# work on your computer

# Note: Because the data in the Crime Open Database can be suject to change, 
# running the code below *could* result in a different data.frame than used 
# in this vignette. Therefore, the resulting datasets are included 
# with the package NearRepeat. I.e., the below code need not be executed: 
# after loading the package you can simply access the data object 'chicago_be'
# and 'chicago_arson'.

library(crimedata)
chicago_df <- get_crime_data(years = 2016, cities = "Chicago", type = "core")
chicago_df <- chicago_df[which(chicago_df$offense_type %in% c("residential burglary/breaking & entering", "arson")), ]

# convert to SpatialPointsDataFrame
library(sp)
library(rgdal)
wgs84_CRS <- CRS("+init=epsg:4326")

chicago_sp <- SpatialPointsDataFrame(coords = cbind(chicago_df$longitude, chicago_df$latitude),
                                     data = chicago_df[, c("uid", "date_single", "offense_type")],
                                     proj4string = wgs84_CRS)

# Re-project spatial data to EPSG:26971 projection (Illinois State Plane)
illinois_CRS <- CRS("+init=epsg:26971")
chicago_sp <- spTransform(chicago_sp, illinois_CRS)

# add X and Y variables, and date
chicago_sp$X = coordinates(chicago_sp)[,1]
chicago_sp$Y = coordinates(chicago_sp)[,2]
chicago_sp$date <- as.Date(chicago_sp$date_single)

chicago_df <- as.data.frame(chicago_sp@data)

To be sure that the NRC software accepts all values, a new dates column is added that uses the "mm/dd/yyyy" format:

chicago_df$dates <- paste(substr(as.character(chicago_df$date), 6, 7),
                          substr(as.character(chicago_df$date), 9, 10),
                          substr(as.character(chicago_df$date), 1, 4),
                          sep = "/")

The data.frame is split up into residential burglary (breaking & entering) and arson, and saves as .csv files.

# save as data.frame
chicago_arson <- chicago_df[which(chicago_df$offense_type == "arson"), c("X", "Y", "date")]

# save as data.frame
chicago_be <- chicago_df[which(chicago_df$offense_type == "residential burglary/breaking & entering"), c("X", "Y", "date")]

# Select first month of data, and save as .csv file
# Adjust file name (and include path) accordingly
chicago_be_month1 <- chicago_be[which(chicago_be$date < "2016-02-01"), ]

write.table(chicago_be_month1,
            file = "inst/extdata/chicago_be_month1.csv",
            sep = ",",
            quote = FALSE,
            row.names = FALSE, col.names = FALSE)

write.table(chicago_arson,
            file = "inst/extdata/chicago_arson.csv",
            sep = ",",
            quote = FALSE,
            row.names = FALSE, col.names = FALSE)

usethis::use_data(chicago_be, overwrite = TRUE)
usethis::use_data(chicago_arson, overwrite = TRUE)

[^1]: Ashby, M.P.J. (2018). Crime Open Database (CODE). doi: 10.17605/OSF.IO/ZYAQN

wsteenbeek/NearRepeat documentation built on Oct. 13, 2020, 9:49 a.m.