basics.R

# load packages ---------

library(tidyverse)
library(here)
library(skimr)
library(janitor)

# read in data -----------

beaches <- read_csv(here("data","sydneybeaches.csv"))

# exploring the data --------------

View(beaches)
dim(beaches)
str(beaches)
glimpse(beaches)
head(beaches)
tail(beaches)
summary(beaches)
skim(beaches)


# tidying columns -------------------

glimpse(beaches)

select_all(beaches, toupper)
select_all(beaches, tolower)

cleanbeaches <- clean_names(beaches)

names(cleanbeaches)

# for rename use newname = oldname
cleanbeaches <- rename(cleanbeaches, beachbugs = enterococci_cfu_100ml)

# select a subset of columns 
select(cleanbeaches, council, site, beachbugs, everything())

# pipe %>%

cleanbeaches <- beaches %>%
  clean_names() %>% 
  rename(beachbugs = enterococci_cfu_100ml) 

write_csv(cleanbeaches, "cleanbeaches.csv")

# sorting and filtering ------------

# which beach has the most extreme levels of bugs

worstbugs <- cleanbeaches %>% arrange(-beachbugs)

worstcoogee <- cleanbeaches %>% 
  filter(site == "Coogee Beach") %>%
  arrange(-beachbugs)

# lets compare max bug values across different beaches

cleanbeaches %>%
  filter(site %in% c("Coogee Beach", "Bondi Beach")) %>%
  arrange(desc(beachbugs))

# group_by and summarise ------------
cleanbeaches %>%
  group_by(site) %>%
  summarise(maxbug = max(beachbugs, na.rm = TRUE), 
            meanbugs = mean(beachbugs, na.rm = TRUE),
            medianbugs = median(beachbugs, na.rm = TRUE),
            sdbugs = sd(beachbugs, na.rm = TRUE))

# lets compare councils

cleanbeaches %>% distinct(council)

councilbysite <- cleanbeaches %>%
  group_by(council, site) %>%
  summarise(meanbugs = mean(beachbugs, na.rm = TRUE), 
            medianbugs = median(beachbugs, na.rm = TRUE)) 



# compute new variables ------------

glimpse(cleanbeaches)

testdate <- cleanbeaches %>% separate(date, c("day", "month", "year"), remove = FALSE)

cleanbeaches %>% unite(council_site, council:site, remove = FALSE)

# use mutate to transform the beachbugs data

summary(cleanbeaches)

cleanbeaches %>% mutate(logbeachbugs = log(beachbugs))

#use mutate to computer new numeric variable

cleanbeaches %>% mutate(beachbugsdiff = beachbugs - lag(beachbugs))

#use mutate to compute new logical variable

cleanbeaches %>% mutate(buggier = beachbugs > mean(beachbugs, na.rm = TRUE))

meanbugs = mean(cleanbeaches$beachbugs, na.rm= TRUE)

cleanbeaches_new <- cleanbeaches %>%
  separate(date, c("day", "month", "year"), remove = FALSE) %>%
  mutate(logbeachbugs = log(beachbugs)) %>%
  mutate(beachbugsdiff = beachbugs - lag(beachbugs)) %>%
  mutate(buggier_all = beachbugs > mean(beachbugs, na.rm= TRUE)) %>%
  group_by(site) %>%
  mutate(buggier_site = beachbugs > mean(beachbugs, na.rm= TRUE))

# write cleaned data to .csv -------------

write_csv(cleanbeaches_new, here("data", "cleanbeaches_new.csv"))