2015-06-08
BP Oil Spill
Data Sets
- NOAA Data
- National Oceanic and Atmospheric Administration
- Temperature and Salinity Data in Gulf of Mexico
- Measured using Floats, Gliders and Boats
- US Fisheries and Wildlife Data
- Animal Sightings on the Gulf Coast
- Birds, Turtles and Mammals
- Status: Oil Covered or Not
Both data sets have geographic coordinates for ever observation
Loading NOAA Data
x <- ls()
noaa <- "http://www.public.iastate.edu/~hofmann/looking-at-data/data/noaa.rdata"
if (!file.exists("noaa.rdata")) download.file(noaa, "noaa.rdata")
load("noaa.rdata")
setdiff(ls(), x)
## [1] "boats" "floats" "gliders" "noaa" "rig" "x"
Floats Data
str(floats)
## 'data.frame': 10332 obs. of 14 variables: ## $ callSign : Factor w/ 11 levels "Q4901043","Q4901044",..: 1 1 1 1 1 1 1 1 1 1 ... ## $ Date_Time : Factor w/ 73 levels "5/24/2010","5/26/2010",..: 30 30 30 30 30 30 30 30 30 30 ... ## $ JulianDay : num 2455390 2455390 2455390 2455390 2455390 ... ## $ Time_QC : int 1 1 1 1 1 1 1 1 1 1 ... ## $ Latitude : num 24.8 24.8 24.8 24.8 24.8 ... ## $ Longitude : num -88 -88 -88 -88 -88 ... ## $ Position_QC : int 1 1 1 1 1 1 1 1 1 1 ... ## $ Depth : int 2 4 6 8 10 12 14 16 18 20 ... ## $ Depth_QC : int 1 1 1 1 1 1 1 1 1 1 ... ## $ Temperature : num 29.8 29.6 29.5 29.5 29.5 ... ## $ Temperature_QC: int 1 1 1 1 1 1 1 1 1 1 ... ## $ Salinity : num 36.6 36.6 36.6 36.6 36.6 ... ## $ Salinity_QC : int 1 1 1 1 1 1 1 1 1 1 ... ## $ Type : Factor w/ 1 level "Float": 1 1 1 1 1 1 1 1 1 1 ...
Floats
library(ggplot2) qplot(Longitude, Latitude, colour = callSign, data = floats) + coord_map()
A note on qplot() versus ggplot()
qplot()makes many assumptions to save you some typing.- Although it's often useful for quick, interactive, exploratory data analysis; eventually you'll want the full flexibility of
ggplot(). qplot()assumes variable names are aesthetics (a mapping from data to visual elements), but withggplot(), you have to specify them.
ggplot(data = floats, aes(x = Longitude, y = Latitude, colour = callSign)) + geom_point() + coord_map()
- RStudio's ggplot2 cheatsheet provides a great visual of this difference.
Gliders Data
str(gliders)
## 'data.frame': 369384 obs. of 14 variables: ## $ callSign : Factor w/ 10 levels "48900","48901",..: 2 2 2 2 2 2 2 2 2 2 ... ## $ Date_Time : Factor w/ 74 levels "5/28/2010","5/29/2010",..: 16 16 16 16 16 16 16 16 16 16 ... ## $ JulianDay : num 2455351 2455351 2455351 2455351 2455351 ... ## $ Time_QC : int 1 1 1 1 1 1 1 1 1 1 ... ## $ Latitude : num 27.9 27.9 27.9 27.9 27.9 ... ## $ Longitude : num -84 -84 -84 -84 -84 ... ## $ Position_QC : int 1 1 1 1 1 1 1 1 1 1 ... ## $ Depth : int 3 4 5 6 7 8 9 10 11 12 ... ## $ Depth_QC : int 1 1 1 1 1 1 1 1 1 1 ... ## $ Temperature : num 28.1 27.8 27.8 27.7 27.7 ... ## $ Temperature_QC: int 1 1 1 1 1 1 1 1 1 1 ... ## $ Salinity : num 35.4 35.4 35.4 35.3 35.3 ... ## $ Salinity_QC : int 1 1 1 1 1 1 1 1 1 1 ... ## $ Type : Factor w/ 1 level "Glider": 1 1 1 1 1 1 1 1 1 1 ...
Gliders
ggplot(data = gliders, aes(x = Longitude, y = Latitude, colour = callSign)) + geom_point() + coord_map()
Boats Data
str(boats)
## 'data.frame': 106735 obs. of 14 variables: ## $ callSign : Factor w/ 2 levels "WTEO","WTER": 1 1 1 1 1 1 1 1 1 1 ... ## $ Date_Time : Factor w/ 43 levels "4/22/2010","4/23/2010",..: 12 12 12 12 12 12 12 12 12 12 ... ## $ JulianDay : num 2455330 2455330 2455330 2455330 2455330 ... ## $ Time_QC : int 1 1 1 1 1 1 1 1 1 1 ... ## $ Latitude : num 26.3 26.3 26.3 26.3 26.3 ... ## $ Longitude : num -87 -87 -87 -87 -87 ... ## $ Position_QC : int 1 1 1 1 1 1 1 1 1 1 ... ## $ Depth : int 2 3 4 5 6 7 8 9 10 11 ... ## $ Depth_QC : int 1 1 1 1 1 1 1 1 1 1 ... ## $ Temperature : num 27.8 27.8 27.7 27.7 27.7 ... ## $ Temperature_QC: int 1 1 1 1 1 1 1 1 1 1 ... ## $ Salinity : num 36.4 36.4 36.4 36.4 36.4 ... ## $ Salinity_QC : int 1 1 1 1 1 1 1 1 1 1 ... ## $ Type : Factor w/ 1 level "Boat": 1 1 1 1 1 1 1 1 1 1 ...
Boats
ggplot(data = boats, aes(x = Longitude, y = Latitude, colour = callSign)) + geom_point() + coord_map()
Provide some context
- All of this data comes from the same geographic region.
- It'd be nice if we could overlay this information onto a map!
- This is where ggplot2's layering idea becomes useful.
states <- map_data("state")
map_outline <- ggplot() +
geom_path(data = states, aes(x = long, y = lat, group = group)) +
xlim(c(-91, -80)) + ylim(c(22,32)) + coord_map()
map_outline
map_floats <- map_outline + geom_point(data = floats, aes(x = Longitude, y = Latitude, colour = callSign)) map_floats
rig # location of BP Oil rig
## x y ## 1 -88.366 28.73663
map_floats +
geom_point(data = rig, aes(x, y), shape = "x", size = 5) +
geom_text(data = rig, aes(x, y), label = "BP Oil rig",
size = 5, hjust = -0.1)
Your Turn
- Use your ggplot2 skills to explore any (or all!) of the floats/gliders/boats data. Be creative!!
How exactly does layering work?
A simpler example
p <- ggplot(data = diamonds, aes(x = carat, y = price)) + geom_point() + # layer 1 geom_smooth(method = lm) # layer 2 p
Plot-level things
# scales p <- p + scale_x_log10() + scale_y_log10() p
# coordinate system p + coord_polar()
# facets p + facet_grid(. ~ cut)
- Each
geom_*function is really a layer with certain defaults for statistic, position, and, well, geometry.
args(geom_point)
## function (mapping = NULL, data = NULL, stat = "identity", position = "identity", ## na.rm = FALSE, ...) ## NULL
args(geom_smooth)
## function (mapping = NULL, data = NULL, stat = "smooth", position = "identity", ## ...) ## NULL
args(geom_bar)
## function (mapping = NULL, data = NULL, stat = "bin", position = "stack", ## ...) ## NULL
Defaults aren't always right!!
base <- ggplot(diamonds, aes(x = cut, fill = clarity)) + theme(legend.position = "none") base + geom_bar() # bad!!! base + geom_bar(position = "fill") # better base + geom_bar(position = "dodge") + theme(legend.position = "bottom") # better
Your Turn
- Read in the animal.csv data
animals <- read.csv("http://heike.github.io/rwrks/02-r-graphics/data/animal.csv")
- Plot the location of animal sightings on a map of the region
- On this plot, try to color points by class of animal and/or status of animal
- Advanced: Could we indicate time somehow?