I have many GPS points and what I want is the sum of distances between two subsequent points (rows) in a given date so I can have a daily track distance.
Each day has about 200 GPS points. Subsequent points means two rows in which the first is earlier than the second. Because I need the total distance among these points, it must take into consideration the row order of the column "time" within a given day (column "date").
Thank you!
My table sorta looks like this:
date time lat lon
18-Jan-18 12:48:39 -24.061464 -47.99523
18-Jan-18 12:48:48 -24.06163 -47.995354
18-Jan-18 12:53:17 -24.06175 -47.995277
We can use pointDistance from the package raster to calculate the distance. lag from dplyr will help calculating on subsequent points. replace_na from tidyr is very convenient, but you could use your favorite way of dealing with NA.
library(raster)
library(dplyr)
library(tidyr)
data %>%
mutate(Distance = pointDistance(cbind(lon,lat),cbind(lag(lon),lag(lat)),lonlat = TRUE)) %>%
mutate(TotalDistance = cumsum(replace_na(Distance,0)))
# date time lat lon Distance TotalDistance
#1 18-Jan-18 12:48:39 -24.06146 -47.99523 NA 0.00000
#2 18-Jan-18 12:48:48 -24.06163 -47.99535 22.29547 22.29547
#3 18-Jan-18 12:53:17 -24.06175 -47.99528 15.42660 37.72207
Data
data <- structure(list(date = structure(c(1L, 1L, 1L), .Label = "18-Jan-18", class = "factor"),
time = structure(1:3, .Label = c("12:48:39", "12:48:48",
"12:53:17"), class = "factor"), lat = c(-24.061464, -24.06163,
-24.06175), lon = c(-47.99523, -47.995354, -47.995277)), class = "data.frame", row.names = c(NA,
-3L))