Function and code for comparing drive times from multiple origins to multiple destinations; with charting

Multi point drive times and compare.r

# Driving distance from multuple starts to multilpe destinations
# modified from the code of gmapsdistance package  by Rodrigo Azuero Melo
# https://cran.r-project.org/web/packages/gmapsdistance

gdist2 <- function (origin, destination, mode, key) 
{
  typeKEY = class(key)
  if (typeKEY != "character") {
    stop("Key should be string")
  }
  if (mode != "driving" & mode != "walking" & mode != "bicycling" & 
        mode != "transit") {
    stop("Mode of transportation not recognized. Mode should be one of 'bicycling', 'transit', 'driving', 'walking' ")
  }
  url0 = "https://maps.googleapis.com/maps/api/distancematrix/xml?origins="
  modeT = paste0("|&mode=", mode, "&language=en-EN")
  urlfin = paste0(url0, origin, "|&destinations=", destination, 
                  modeT, "&key=", key)
  # urlfin <- "https://maps.googleapis.com/maps/api/distancematrix/xml?origins=Flourtown+PA|Philadelphia+PA&destinations=Burlington+NJ|&mode=driving&language=en-EN&key=AIzaSyBCkmICVDoIkV055x6l-gM-1Le82A3VsEI"
  webpageXML <- getURL(urlfin)
  webpageXML2 <- readLines(tc <- textConnection(webpageXML))
  close(tc)
  webpageXMLA <- xmlParse(webpageXML2)
  root = xmlRoot(webpageXMLA)
  first = xmlChildren(root)
  
  # for multi origin
  initstatus = first$status[[1]]
  initstat = as(initstatus, "character")
  length_origin <- strsplit(origin,"|",fixed=TRUE)
  length_origin <- length(length_origin[[1]])
  origins_xml <- first[c(2:(length_origin+1))]
  origins <- NULL
  for(i in 1:length_origin){
    origin_parse <- as(origins_xml[[i]][[1]], "character")
    origin_parse <- strsplit(origin_parse,",")[[1]][1]
    origins <- c(origins, origin_parse)
  }
  dest <- as(first[[length_origin+2]][[1]], "character")
  rows = first[c((length_origin+3):(length_origin*2+3))]
  TDS <- NULL
  for(i in 1:length_origin){
    row_xml <- rows[[i]]
    Status = xmlChildren(row_xml[[1]])
    Status = Status$status[1]
    Status = Status$text
    Status = as(Status, "character")
    Time = xmlChildren(row_xml[[1]])
    Time = Time$duration[1]
    Time = Time$value[1]$text
    Time = round(as(Time, "numeric")/60,1)
    Distance = xmlChildren(row_xml[[1]])
    Distance = Distance$distance[1]
    Distance = Distance$value[1]$text
    Distance = round(as(Distance, "numeric")*0.000621371,1) # convert meter to mile
    MPH <- round(Distance/(Time/60),2)
    output = data.frame(Time, Distance, MPH, Status)
    TDS <- rbind(TDS, output)
  }
  TDS <- cbind(origins, TDS)
  return(TDS)
}

require(RCurl)
require(XML)
require(ggplot2)

key <- "YOUR KEY HERE!!!"
mode <- "driving"
towns <- c("Wayne", "Plymouth+Meeting", "Germantown", "Flourtown","Oreland",
           "Phoenixville", "Ambler", "Horsham", "King+of+Prusia", "Royersford",
           "Erdehiem", "Wyndmoor", "Lafayette+Hill", "Ardmore", "Willowgrove",
           "Blue+Bell", "Trooper", "Audubon", "Trappe", "Skippack", "Collegeville",
           "Landsdale", "North+Whales", "Pottstown", "Fort+Washington", "East+Norriton")
origin <- paste(towns, collapse="+PA|")
origin <- paste(c(origin, "+PA"), collapse = "")
destination1 <- "Burlington+NJ"
destination2 <- "Conshohocken+PA"

results_burl <- gdist2(origin, destination1, mode, key)
results_conchy <- gdist2(origin, destination2, mode, key)

both_locals <- cbind(results_burl[,1:4], results_conchy[,1:4])
both_locals$joint <- (both_locals[,2] + both_locals[,6])/2
both_locals <- (both_locals[,c(1,2,3,4,6,7,8,9)])
colnames(both_locals) <- c("Origin", "Burl-Time", "Burl-Miles", "Burl-MPH",
                           "Conchy-Time", "Conchy-Miles", "Conchy-MPH",
                           "Avg-Time")
both_locals$`Weighted-Avg` <- apply(both_locals[,c("Burl-Time","Conchy-Time")],1,
                          weighted.mean,w=c(0.65,0.35))
both_locals <- both_locals[order(both_locals$`Weighted-Avg`),]
bcost <- both_locals$`Burl-MPH`
print(both_locals)

ggplot(both_locals, aes(x = Origin, y = `Weighted-Avg`, fill = `Weighted-Avg`)) +
  geom_bar(stat="identity") +
  scale_x_discrete(limits=rev(both_locals$Origin)) +
  geom_hline(yintercept = both_locals[which(both_locals$Origin == "Germantown"),"Weighted-Avg"],
             size = 1, color = "red", linetype="dashed") +
  coord_flip()