1 A short Introduction to the Vignette

I’m very new in the R world. Welcome to my journey of learning Geospatial visualisation in R :)

The purpose of this vignette is to visualise how many blackspot mobile sites were awarded in QLD on map, and calculate the distance between each site and Brisbane General Post Office(GPO). As blackspot sites are normally very far from the capital city, we need to know the travel distance in order to determine LAFA(Living away from home allowance) expenses for our field team. This is a very useful tool when it comes to cost tracking/forecasting budget for rural build projects. I downloaded the Mobile Black Spot Program (MBSP)dataset on QLD Gov website(source 1).

2 Install and load packages

We are using ggmap to visaliase the mobile base stations locations on map. We also need to register a free Google API key and enable relevant API tools in the library. I will not go into details of API keys as this is not the main purpose of this Vignette. Please refer to the instruction on how to get API Key https://developers.google.com/maps/documentation/javascript/get-api-key (source 2)

library(tidyverse)
library(dplyr)
library(ggmap)
register_google(key = "AIzaSyB1Fccl9sZOEUWC0NISGPdW5pclAp6vyo0")
library(knitr)
library(leaflet)
library(gmapsdistance)

3 Read and check datasets from downloaded CSV file.

We can usekable from knitrpackage to tidy up the table. Usehead()function to check the first 6 rows of the dataset.

blackspot<-read.csv(file='QLD blackspot sites.csv')
knitr::kable(head(blackspot[1:6, ]), caption ="QLD Mobile blackspot program")
QLD Mobile blackspot program
LGA BASE_STATION_ID LOCATION_NAME TITLE ROAD_SECTION MOBILE_OPERATOR BASE_STATION_TYPE STATE_ELECTORATE Latitude Longitude ROUND
Balonne OPT-007 Castlereagh Highway Castlereagh Highway Between St George and Dirranbandi (Blue Lagoon) Optus Macro tower WARREGO -28.32323 148.4792 2
Balonne TLS-1114 Bundoran Road Bundoran Road N/A Telstra Macro tower WARREGO -28.11330 148.7275 2
Barcaldine OPT-029 Landsborough Highway Landsborough Highway Avoca Station Optus Macro tower GREGORY -23.53730 144.7930 2
Blackall-Tambo TLS-0544 Landsborough Highway Landsborough Highway Telstra Macro tower GREGORY -25.34400 146.5064 2
Brisbane TLS-0492 Kholo Kholo N/A Telstra Macro tower MOGGILL -27.50506 152.7749 2
Brisbane TLS-1723 Upper Brookfield Upper Brookfield N/A Telstra Macro tower MOGGILL -27.47553 152.8692 2

4 Get QLD map and visulise the sites

First, we need to define an object: qldmap.Then, we use gggmap()``geom_point()to plot the sites on the map. x=longitude and y=latitude . As you can see in the map below, we have some outliers(sites located on Cook island) and some overplotting issues.Instead of generating static maps, Can we plot an interactive map with zoom-in and zoom-out functions?

qldmap<-get_map(location = "QLD,Australia", zoom=5, source = "google")
ggmap(qldmap)+
geom_point(aes(Longitude,Latitude,colour=BASE_STATION_TYPE,shape=MOBILE_OPERATOR), data=blackspot)

5 Interactive map - Zoom in and Zoom out

We have already installed the leaflet package and loaded the library. Create map by callingleaflet() and add layers like addTiles() and addMarkers. Tiles are not specified in below code so OpenStreet map will be used by default. Now we can zoom in and zoom out to check every site name.

leaflet(data=blackspot) %>% 
  addTiles() %>% ##default map= openstreet map
  addMarkers(~Longitude, ~Latitude,popup=~LOCATION_NAME) ##drop pins on map, click to see location names

6 Now calculate distance and travel time

To calculate LAFA(Living away from Home allowance) expense, we normally will set the General Post office in each Capital city as our starting point, and destinations are the QLD black spots sites.

Now we need to get distance and travel time between Brisbane GPO and mobile sites. Use gmapsdistance(from,to, mode= "driving"...) (Source 3)to get the time, distance and status. For example, if we set the site on Cook Island as our destination, “status” will be invalid as gmapsdistance() cannot calculate distance over ocean. You can also choose “walking”,“bicycling” and “public transportation” as traveling mode.

I also create a dataframe Timedist2 to store distance/travel time for each site.

set.api.key("AIzaSyB1Fccl9sZOEUWC0NISGPdW5pclAp6vyo0") ## enable Google Distance Matrix API and set API key

mobilesites=paste(blackspot$Latitude,blackspot$Longitude,sep =",")
GPO = "-27.467676,153.027908" ## Brisbane General post office Coordinates

Timedist=gmapsdistance(GPO, mobilesites, mode="driving", shape = "long")
Timedist2=data.frame(Timedist) ##create dataframe
names(Timedist2) <- c("GPO","Destination", "Time", "Location_Name","col2","Distance","col3","col4","Status")

Timedist2$Location_Name <-blackspot$LOCATION_NAME ## Add location name
Timedist2$col2 <- NULL ## delete repeated cols        
Timedist2$col3<- NULL  
Timedist2$col4 <- NULL

Timedist2$Time <- Timedist2$Time/3600 ##  hour
Timedist$Distance <- Timedist2$Distance/1000 ## KM

knitr::kable(head(Timedist2[1:6,]),caption ="QLD Mobile Blackspots sites-Time and distance")
QLD Mobile Blackspots sites-Time and distance
GPO Destination Time Location_Name Distance Status
-27.467676,153.027908 -28.32323337,148.479157 5.9661111 Castlereagh Highway 539649 OK
-27.467676,153.027908 -28.113298,148.72747 5.6075000 Bundoran Road 508413 OK
-27.467676,153.027908 -23.5373,144.793 12.1361111 Landsborough Highway 1122969 OK
-27.467676,153.027908 -25.344001,146.506428 8.8200000 Landsborough Highway 800815 OK
-27.467676,153.027908 -27.505065,152.774925 0.7763889 Kholo 36091 OK
-27.467676,153.027908 -27.475527,152.869202 0.5269444 Upper Brookfield 20464 OK

I really enjoyed doing my first R assignment. Please feel free to leave any comments :D

Source:

1.https://www.data.qld.gov.au/dataset/qld-base-stations-mobile-black-spot-program

2.https://developers.google.com/maps/documentation/javascript/get-api-key

3.https://cran.r-project.org/web/packages/gmapsdistance/gmapsdistance.pdf