This is a further question from how to create data frames (not just one) at once in r. I really can't understand the error message. I have two data frames:
> route1
X1 X2 X3
1 Balboa Park, san diego, ca Coronado, san diego, ca USS Midway, san diego, ca
2 Balboa Park, san diego, ca USS Midway, san diego, ca Coronado, san diego, ca
3 Coronado, san diego, ca Balboa Park, san diego, ca USS Midway, san diego, ca
4 Coronado, san diego, ca USS Midway, san diego, ca Balboa Park, san diego, ca
5 USS Midway, san diego, ca Balboa Park, san diego, ca Coronado, san diego, ca
6 USS Midway, san diego, ca Coronado, san diego, ca Balboa Park, san diego, ca
and
> place
# A tibble: 8 x 5
name hour type short abbr
* <chr> <dbl> <chr> <chr> <chr>
1 Balboa Park, san diego, ca 4 place Balboa Park bal
2 USS Midway, san diego, ca 2 place USS Midway midway
3 Coronado, san diego, ca 4 place Coronado coro
4 Cabrillo National Monument, san diego, ca 2 place Cabrillo cab
5 La Jolla Cove, san diego, ca 2 place La Jolla ljc
6 4002 Wallace St, San Diego, CA 92110 4 place Old Town ot
7 Hotel Republic San Diego, Autograph Collection NA hotel Autograph Collection NA
8 THE US GRANT, a Luxury Collection Hotel, San Diego NA hotel Luxury Collection NA
My idea is using ggmap package to get driving time from Hotel Republic San Diego, Autograph Collection (place$name[7]) to the first place in route1's first row, then get driving time from the first place in route1's first row to the second place in route1's first row, and so on, until from the last place in route1's first row to THE US GRANT, a Luxury Collection Hotel, San Diego(place$name[8]). Sum up the time, store in route1$time column. Then do the same thing for the following rows. So my code is
library(ggmap)
register_google(key = "my_google_key_here")
route1$time <- 0
for (i in 1:length(route1)) {
for (j in 1:length(route1[i,])) {
if(j == 1){
route1$time[i] <- route1$time[i] + mapdist(from = place$name[7], to = as.character(route1[i,][1]))$seconds
}
if(j == length(route1[i,])){
route1$time[i] <- route1$time[i] + mapdist(from = as.character(route1[i,][j]), to = place$name[8])$seconds
route1$time[i] <- route1$time[i] + mapdist(from = as.character(route1[i,][j - 1]), to = as.character(route1[i,][j]))$seconds
}
for (j in 2:length(route1[i,]-1)) {
route1$time[i] <- route1$time[i] + mapdist(from = as.character(route1[i,][j - 1]), to = as.character(route1[i,][j]))$seconds
}
}
}
Executing this code gives me
Error in FUN(left, right) : non-numeric argument to binary operator
error message, I think it is saying that I add a character to a number, but I can't find it.
# this is a nice library for R
library(tidyverse)
# name your variables
# (you can use these as arguments to a function or something)
start = place$name[7]
end = place$name[8]
# wrap your computation into a function which takes
# a row of a data frame as input
compute_route_time = function(row) {
# get a vector of stops that you can loop through
stops = unlist(row)
# distance from start to first stop
time = mapdist(from = start, to = stops[1])$seconds
# iterate through the rest of the route
n_stops = length(stops)
for (i in 1:(n_stops-1)) {
time = time + mapdist(from = stops[i], to = stops[i+1])$seconds
}
# finish rout
time = time + mapdist(from = stops[n_stops], to = end)$seconds
# collect row as a data frame to preserve old columns
row = as.data.frame(row)
# add time column
row$time = time
return(row)
}
route1 %>%
dplyr::rowwise() %>%
dplyr::do(compute_route_time(.))