I'm plotting a map, which is supposed to have an overlay of several (> 400) arrows from a dataset, which has lat/lon pairs for both start and end of each arrow. Here is a subset of the data using dput:
df <- structure(list(Lat = c(49.34054, 49.34068, 49.3409, 49.34106,
49.34116, 49.34133, 49.34138, 49.34144, 49.34155, 49.34164, 49.34168,
49.34178, 49.34179, 49.34187, 49.34199, 49.34202, 49.3421, 49.34219,
49.34226, 49.34236, 49.3424), Lon = c(-117.76365, -117.76433,
-117.76474, -117.76575, -117.76646, -117.76607, -117.76643, -117.76676,
-117.76611, -117.76638, -117.76678, -117.76612, -117.7671, -117.76678,
-117.76776, -117.76745, -117.76706, -117.76815, -117.76778, -117.76762,
-117.76812), LatEnd = c(49.3404216917208, 49.3404813977525, 49.3407696999527,
49.3409218055133, 49.3408834255181, 49.3411571438575, 49.3411444104952,
49.3412068979592, 49.3413453850968, 49.3414853385912, 49.3414067819334,
49.3415646398153, 49.3415782191525, 49.3416671210859, 49.341769715577,
49.3418702688525, 49.3418749917805, 49.3419107724121, 49.3418905356976,
49.3421097403974, 49.3419881060831), LonEnd = c(-117.76319214364,
-117.763951728004, -117.76423214535, -117.765201260725, -117.766005995062,
-117.765592930919, -117.765831425366, -117.766367701412, -117.765558298351,
-117.765915748408, -117.766324336458, -117.765721708226, -117.766709104693,
-117.766420637063, -117.767340559198, -117.767000983228, -117.766699658212,
-117.767827633167, -117.767235785716, -117.767302080608, -117.767699155441
)), .Names = c("Lat", "Lon", "LatEnd", "LonEnd"), row.names = c(244L,
263L, 293L, 313L, 330L, 351L, 359L, 369L, 390L, 409L, 414L, 426L,
427L, 435L, 442L, 443L, 448L, 450L, 455L, 457L, 459L), class = "data.frame")
I've worked with ggmap before, and therefore this was my first try:
library(ggplot2)
library(ggmap)
prep <- get_googlemap(
center = c(-117.7670, 49.34027), #Long/lat of centre, or "Edinburgh"
zoom = 17,
maptype = 'hybrid', #also hybrid/terrain/roadmap/satellite
scale = 2)
map <- ggmap(prep,
size = c(100, 200),
extent='device',
darken = 0.5,
legend = "bottom",
base_layer = ggplot(data = df, aes(x = Lon, y = Lat)))
p <- map +
geom_segment(aes(xend = LonEnd, yend = LatEnd),
arrow = arrow(length = unit(0.2,"cm"), angle = 15), size = 0.3, colour = "white")
Which produced a very odd result - some of the arrows had their heads plotted in the wrong direction.
Then I proceeded to try and use geom_path instead. Which corrected for the direction of the heads, but required that I input every line of the dataset separately as input to the geom_path. I made a for loop, but apparently what happened there was that every time a layer was added ('i' upped by one), all previous layers would disappear.
Lons <- cbind(df$Lon, df$LonEnd)
Lats <- cbind(df$Lat, df$LatEnd)
map <- ggmap(prep,
size = c(100, 200),
extent='device',
darken = 0.5,
legend = "bottom",
base_layer = ggplot(data = df[1,], aes(x = Lon, y = Lat)))
p <- map + geom_path(aes(x = Lons[10,], y = Lats[10,]),
arrow = arrow(length = unit(0.2,"cm"), angle = 15), size = 0.3, colour = "white")
for(i in 2:nrow(Lats)){
p <- p + geom_path(aes(x = Lons[i,], y = Lats[i,]),
arrow = arrow(length = unit(0.2,"cm"), angle = 15), size = 0.3, colour = "white")
} #i
Any help would be most appreciated...