Working with Categorical Data in R

forcats.R

# install
install.packages('forcats')
install.packages('readr')
install.packages('dplyr')
install.packages('ggplot2')

# library
library(forcats)
library(readr)
library(dplyr)
library(ggplot2)

# import data
ecom <- readr::read_csv('https://raw.githubusercontent.com/rsquaredacademy/datasets/master/web.csv')
ecom

# tabulate referrers
ecom %>%
  count(referrer)

# average page visits by referrers
refer_summary <- ecom %>%
  group_by(referrer) %>%
  summarise(
    page = mean(n_pages),
    tos = mean(duration),
    n = n()
  )

ggplot(refer_summary, aes(page, referrer)) + geom_point()

ggplot(refer_summary, aes(page, fct_reorder(referrer, page))) + geom_point()

# referrer frequency
ecom %>%
  mutate(ref = referrer %>% fct_infreq()) %>%
  ggplot(aes(ref)) +
  geom_bar()

ecom %>%
  mutate(ref = referrer %>% fct_infreq() %>% fct_rev()) %>%
  ggplot(aes(ref)) +
  geom_bar()

# import data
traffic <- readr::read_csv('https://raw.githubusercontent.com/rsquaredacademy/datasets/master/web_traffic.csv')
traffic

# tabulate referrer
traffic$traffics %>% 
  fct_count()

# collapse referrer categories
traffic2 <- fct_collapse(traffic$traffics,
  social = c("facebook", "twitter", "instagram"),
  search = c("google", "bing", "yahoo")
)

traffic2 %>% fct_count() 

# lump infrequent referrers
traffic$traffics %>% 
  fct_lump() %>% 
  table()

# retain top 3 referrers
traffic$traffics %>% 
  fct_lump(n = 3) %>% 
  table()

# lump together referrers with < 10% traffic
traffic$traffics %>% 
  fct_lump(prop = 0.1) %>% 
  table()

# lump together referrers with < 15% traffic
traffic$traffics %>% 
  fct_lump(prop = 0.15) %>% 
  table()

# retain 3 referrers with lowest traffic
traffic$traffics %>% 
  fct_lump(n = -3) %>% 
  table()

# retain 3 referrers with < 10% traffic
traffic$traffics %>% 
  fct_lump(prop = -0.1) %>% 
  table()