{"id":3700,"date":"2015-10-04T20:01:20","date_gmt":"2015-10-05T01:01:20","guid":{"rendered":"http:\/\/rud.is\/b\/?p=3700"},"modified":"2022-06-29T05:38:31","modified_gmt":"2022-06-29T10:38:31","slug":"replicating-natgeos-proper-earthquake-map-in-r","status":"publish","type":"post","link":"https:\/\/rud.is\/b\/2015\/10\/04\/replicating-natgeos-proper-earthquake-map-in-r\/","title":{"rendered":"Replicating NatGeo’s “Proper” Earthquake Map in R"},"content":{"rendered":"

I saw this post<\/a> over at NatGeo over the weekend and felt compelled to replicate this:<\/p>\n

\"\"<\/p>\n

with ggplot2.<\/p>\n

Three shapefiles later<\/a> and we have it close enough to toss into a post (and I really don’t believe the continent names are necessary).<\/p>\n

library(rgdal)\nlibrary(ggplot2)\nlibrary(ggthemes)\nlibrary(ggalt) # devtools::install_github(\"hrbrmstr\/ggalt\")\n\n# grab these from http:\/\/rud.is\/dl\/quakefiles.tgz\n\nworld <- readOGR(\"countries.geo.json\", \"OGRGeoJSON\", stringsAsFactors=FALSE)\nplates <- readOGR(\"plates.json\", \"OGRGeoJSON\", stringsAsFactors=FALSE)\nquakes <- readOGR(\"quakes.json\", \"OGRGeoJSON\", stringsAsFactors=FALSE)\n\nworld_map <- fortify(world)\nplates_map <- fortify(plates)\nquakes_dat <- data.frame(quakes)\nquakes_dat$trans <- quakes_dat$mag %% 5\n\ngg <- ggplot()\ngg <- gg + geom_cartogram(data=world_map, map=world_map,\n                          aes(x=long, y=lat, map_id=id),\n                          color=\"white\", size=0.15, fill=\"#d8d8d6\")\ngg <- gg + geom_cartogram(data=plates_map, map=plates_map,\n                          aes(x=long, y=lat, map_id=id),\n                          color=\"black\", size=0.1, fill=\"#00000000\", alpha=0)\ngg <- gg + geom_point(data=quakes_dat,\n                      aes(x=coords.x1, y=coords.x2, size=trans),\n                      shape=1, alpha=1\/3, color=\"#d47e5d\", fill=\"#00000000\")\ngg <- gg + geom_point(data=subset(quakes_dat, mag>7.5),\n                      aes(x=coords.x1, y=coords.x2, size=trans),\n                      shape=1, alpha=1, color=\"black\", fill=\"#00000000\")\ngg <- gg + geom_text(data=subset(quakes_dat, mag>7.5),\n                     aes(x=coords.x1, y=coords.x2, label=sprintf(\"Mag %2.1f\", mag)),\n                     color=\"black\", size=3, vjust=c(3.9, 3.9, 5), fontface=\"bold\")\ngg <- gg + scale_size(name=\"Magnitude\", trans=\"exp\", labels=c(5:8), range=c(1, 20))\ngg <- gg + coord_map(\"mollweide\")\ngg <- gg + theme_map()\ngg <- gg + theme(legend.position=c(0.05, 0.99))\ngg <- gg + theme(legend.direction=\"horizontal\")\ngg <- gg + theme(legend.key=element_rect(color=\"#00000000\"))\ngg\n<\/code><\/pre>\n
\"unnamed-chunk-1-1\"<\/p>\n
I can only imagine how many mouse clicks that would be in a GIS program.<\/p>\n
Addendum<\/strong><\/p>\n