Skip to content

Latest commit

 

History

History
253 lines (174 loc) · 9.12 KB

visualization.md

File metadata and controls

253 lines (174 loc) · 9.12 KB
Raw
title shorttitle layout lang identifier breadcrumbs
Data visualization using R (deprecated)
Data visualization using R (deprecated)
page
en
visualization
manual

Data visualization using R (deprecated)

Contents

Introduction to the ggplot2 package

For making static maps in R, we are going to use the ggplot2 package. First fetch some Southern sunfish occurrences from OBIS:

library(robis)
molram <- occurrence("Mola ramsayi")

Now create a simple scatter plot using the occurrence coordinates. Use the ggplot() function to initialize a new plot, and geom_point() to create a scatter plot. The aes() function is used to create a mapping (aesthetic) between our data and the visual properties, which is then passed to the geom_point() function. Here decimalLongitude is used as the x coordinate and decimalLatitude as the y coordinate:

library(ggplot2)
ggplot() + geom_point(data = molram, aes(x = decimalLongitude, y = decimalLatitude))

Let's now add a world polygon to our map. ggplot2 provides a map_data() function to load maps of countries, continents or the entire world (this requires the maps package as well). Besides the world polygon, we also pass an aesthetic and a fill parameter to the geom_polygon() function.

library(maps)
world <- map_data("world")

ggplot() +
 geom_polygon(data = world, aes(x = long, y = lat, group = group), fill = "#dddddd") +
 geom_point(data = molram, aes(x = decimalLongitude, y = decimalLatitude))

Let's now use coord_fixed() to make sure the axes have the same scale. You may want to pass a different value to this function if you are mapping areas close to the poles:

ggplot() +
 geom_polygon(data = world, aes(x = long, y = lat, group = group), fill = "#dddddd") +
 geom_point(data = molram, aes(x = decimalLongitude, y = decimalLatitude)) +
 coord_fixed(1)

Let's now zoom in a bit by passing xlim and ylim to coord_fixed:

ggplot() +
 geom_polygon(data = world, aes(x = long, y = lat, group = group), fill = "#dddddd") +
 geom_point(data = molram, aes(x = decimalLongitude, y = decimalLatitude)) +
 coord_fixed(1, xlim = c(0, 180), ylim = c(-60, 0))

To make the plot a bit more interesting, add color to the geom_point aesthetic. In this example the dots are colored based on the institutionCode field of the occurrence data:

ggplot() +
 geom_polygon(data = world, aes(x = long, y = lat, group = group), fill = "#dddddd") +
 geom_point(data = molram, aes(x = decimalLongitude, y = decimalLatitude, color = datasetName)) +
 coord_fixed(1, xlim = c(0, 180), ylim = c(-60, 0))

There are many ways to change color scales in ggplot, look into scale_color_brewer() for example:

ggplot() +
 geom_polygon(data = world, aes(x = long, y = lat, group = group), fill = "#dddddd") +
 geom_point(data = molram, aes(x = decimalLongitude, y = decimalLatitude, color = datasetName)) +
 coord_fixed(1, xlim = c(0, 180), ylim = c(-60, 0)) +
 scale_color_brewer(palette = "Paired")

The geom_histogram() function can be used to create histograms. To try this, first fetch a bit more data from OBIS:

dor <- occurrence("Doridoidea")

Now create a simple histogram:

ggplot() +
 geom_histogram(data = dor, aes(x = yearcollected))

This produces a histogram, but we also get this warning message:

stat_bin() using bins = 30. Pick better value with binwidth. This means that by default there are 30 bins in the histogram. However, because we are displaying records per year, it makes more sense to pick a bin width ourselves, for example 2 years:

ggplot() +
 geom_histogram(data = dor, aes(x = yearcollected), binwidth = 2)

By adding fill to the aesthetic, we can color the bars based on the family:

ggplot() +
 geom_histogram(data = dor, aes(x = yearcollected, fill = family), binwidth = 2) +
 scale_fill_brewer(palette = "Spectral")

Using xlim() we can zoom in a bit:

ggplot() +
 geom_histogram(data = dor, aes(x = yearcollected, fill = family), binwidth = 2) +
 scale_fill_brewer(palette = "Spectral") +
 xlim(c(1950, 2017))

In case you need to split up your graph based on one or more factors, you can use facet_grid(). For example:

library(dplyr)
lag <- occurrence("Lagis")
lag_2 <- lag %>% filter(resourceID %in% c(4312, 222))

ggplot() +
 geom_histogram(data = lag_2, aes(x = yearcollected), binwidth = 2) +
 facet_grid(resourceID ~ species)

Make sure to take a look at the R graph gallery and the ggplot extension gallery and be inspired!

Creating interactive maps using the leaflet package

The leaflet package is a wrapper around the popular Leaflet JavaScript library for interactive maps. Install the package as follows:

install.packages("leaflet")

A simple map

Initialize a map with leaflet() and add the default OpenStreetMap basemap using addTiles():

library(leaflet)

leaflet() %>% addTiles()

To change the basemap, pick any of the tile providers here and pass the URL to addTiles():

leaflet() %>% addTiles("https://server.arcgisonline.com/ArcGIS/rest/services/Ocean_Basemap/MapServer/tile/{z}/{y}/{x}")

Now fetch some data using the robis package, and add circle markers to the map using addCircleMarkers(). This function accepts lng, lat, as well as some styling patameters:

library(robis)
abrseg <- occurrence("Abra segmentum")

leaflet() %>%
  addTiles("https://cartodb-basemaps-{s}.global.ssl.fastly.net/light_all/{z}/{x}/{y}.png") %>%
  addCircleMarkers(lat = abrseg$decimalLatitude, lng = abrseg$decimalLongitude, radius = 3.5, weight = 0, fillOpacity = 1, fillColor = "#cc3300")

Quality flags

The data retrieved using the R package include the OBIS quality flags. The example below visualizes one of these flags for the European sea sturgeon. It also adds popups to the Leaflet map. Use the qcflags() function from the robis package to check for which records flag 28 is set, then use the result to create a vector of colors (red and green).

library(robis)
library(leaflet)

acistu <- occurrence("Acipenser sturio")

acistu$qcnum <- qcflags(acistu$qc, c(28))
colors <- c("#ee3300", "#86b300")[acistu$qcnum + 1]
popup <- paste0(acistu$datasetName, "<br/>", acistu$catalogNumber, "<br/><a href=\"http://www.iobis.org/explore/#/dataset/", acistu$resourceID, "\">OBIS dataset page</a>")

leaflet() %>%
  addProviderTiles("CartoDB.Positron") %>%
  addCircleMarkers(popup = popup, lat = acistu$decimalLatitude, lng = acistu$decimalLongitude, radius = 3.5, weight = 0, fillColor = colors, fillOpacity = 1)

We cann repeat this procedure for multiple quality flags (27 and 29) as follows:

library(robis)
library(leaflet)

ices <- occurrence(resourceid = 1575, enddate = "1985-01-01")

ices$qcnum <- qcflags(ices$qc, c(27, 29))
colors <- c("#ee3300", "#ff9900", "#86b300")[ices$qcnum + 1]

leaflet() %>%
  addTiles("http://{s}.basemaps.cartocdn.com/light_nolabels/{z}/{x}/{y}.png") %>%
  addCircleMarkers(popup = ices$scientificName, lat = ices$decimalLatitude, lng = ices$decimalLongitude, radius = 3.5, weight = 0, fillColor = colors, fillOpacity = 1)

Multiple species

In the example below, data is retrieved and visualized for two cod species.

pac <- occurrence("Gadus macrocephalus")
atl <- occurrence("Gadus morhua", year = 2011)

leaflet() %>%
  addProviderTiles("CartoDB.Positron") %>%
  addCircleMarkers(lat = pac$decimalLatitude, lng = pac$decimalLongitude, radius = 3.5, weight = 0, fillOpacity = 1, fillColor = "#ff0066") %>%
  addCircleMarkers(lat = atl$decimalLatitude, lng = atl$decimalLongitude, radius = 3.5, weight = 0, fillOpacity = 1, fillColor = "#0099cc")