UQRUG 26

2022-04-27: UQRUG 26

Attendees

• Stéphane
• Veronika
• Chris
• Thuong
• Lily
• David

Topics discussed and code

• Machine learning with caret and glmnet
• High-performance computing: https://rcc.uq.edu.au/high-performance-computing
• Spatial data: sf, sfnetworks… Austroad dashboard
• Interactive viusalisations: plotly, highcharter, networkD3, leaflet, tmap, crosstalk, Shiny…
• API / direct link for accessing government data that gets updated weekly (see below)

Tide data

# Whyte island, station measuring tide level

path <- "http://opendata.tmr.qld.gov.au/Whyte_Island.txt"
# read with base function, ignore first lines, keep two columns
tide_data <- read.table(path, skip = 5)[,1:2]
# name the column
names(tide_data) <- c("date_time", "LAT")

# same with readr
library(readr)
library(dplyr)
tide_data <- read_table(path, skip = 5,
                        col_names = FALSE) %>% 
  select(1:2) %>% 
  rename(date_time = 1, LAT = 2)

# split the date time
library(lubridate)
tide_data <- tide_data %>% 
  mutate(date_time = dmy_hm(date_time))
# filter and visualise
library(ggplot2)
tide_data %>% 
  filter(LAT > 0.01) %>% 
  ggplot(aes(x = date_time, y = LAT)) +
  geom_line()

# save only the first time:
# write.csv(tide_data, "all_tide_data.csv", row.names = FALSE)

# append new data
all_tide_data <- read_csv("all_tide_data.csv")
all_tide_data <- bind_rows(all_tide_data, tide_data) %>% 
  unique() # check for duplicates
# overwrite file
write.csv(all_tide_data, "all_tide_data.csv", row.names = FALSE)

Automate running the script (on Windows): https://cran.r-project.org/web/packages/taskscheduleR/index.html