UQRUG 27

meeting
Questions: ggplot2, species distribution modelling, anomalies in chronology, R Markdown, factoextra
Published

June 1, 2022

2022-06-01: UQRUG 27

Attendees

  • Stéphane: Library, here to help!
  • Emma: looking for some help editing a graphic with ggplot2
  • Rene: just tagging along
  • Laura: working on GLMs
  • Leonie: both tagging along and hoping for some help with the adehabitat package
  • Chris: just tagging along
  • David: using R on HPC
  • Astrid: R Markdown issues
  • Olalekan: just tagging along
  • …and 6 other UQRUGers!

Topics discussed and code

  • ggplot2 customisation: moving legend, filtering data out 
library(dplyr)
# do not keep these three eye colours
starwars %>% 
  filter(!eye_color %in% c("blue", "yellow", "green"))
  • Preparation for species distribution modelling. Convert dataframe to sf object with st_as_sf(), and will probably need to go from vector data to raster data with terra::rasterize()
  • Importing spatial points for dolphin occurrences, using sf. Constructing a convex hull from them and visualising on an interactive map:
# read CSV as dataframe
dolph <- read.csv("Adehabitat.csv")

library(sf)
# convert the dataframe to an sf object
dolph_sf <- st_as_sf(dolph, coords = c("Longitude", "Latitude"))
# see it with default plot method
plot(dolph_sf)

# interactive map
library(tmap)
tmap_mode("view")
tm_shape(dolph_sf) +
  tm_dots()

# convex hull
dolph_hull <- st_convex_hull(st_union(dolph_sf))

# visualise both
tm_shape(dolph_hull) +
  tm_borders() +
tm_shape(dolph_sf) +
  tm_dots()
  • Detecting anomalies in chronological sequence of a dataframe. dplyr::lag() and dplyr::lead() functions can be used for comparisons. any() and all() help reducing many logical values to one.
  • R Markdown troubles: Rmd is self-contained and needs to include all the necessary code. Its working directory is by default the directory where the .Rmd file is saved.
  • factoextra’s fviz_pca*() functions for PCA, colouring points per group.