## ----setup, include=FALSE-----------------------------------------------------
library(knitr)
hook_output = knit_hooks$get('output')
knit_hooks$set(output = function(x, options) {
  # this hook is used only when the linewidth option is not NULL
  if (!is.null(n <- options$linewidth)) {
    x = knitr:::split_lines(x)
    # any lines wider than n should be wrapped
    if (any(nchar(x) > n)) x = strwrap(x, width = n)
    x = paste(x, collapse = '\n')
  }
  hook_output(x, options)
})

knitr::opts_chunk$set(cache = FALSE, message = FALSE,
                      linewidth = 50)

library(tidyverse)
glimpse <- purrr::partial(glimpse, width = 50)


## ----message = FALSE----------------------------------------------------------
library(tidyverse)

df_beers <- read_csv("beers.csv")
df_breweries <- read_csv("breweries.csv")

glimpse(df_beers)
glimpse(df_breweries)


## -----------------------------------------------------------------------------
library(tidyverse)

dataset <- inner_join(df_beers,
                      df_breweries,
                      by = "brewery_id")
glimpse(dataset)


## -----------------------------------------------------------------------------
# dataset and df_beers have the same # of rows
nrow(dataset) == nrow(df_beers)
# dataset has one less than the sum of # cols
c(ncol(dataset), ncol(df_beers), ncol(df_breweries))


## -----------------------------------------------------------------------------
# Careful about missing values!
dataset %>% 
  group_by(state) %>% 
  summarise(avg_abv = mean(abv, na.rm = TRUE)) %>% 
  filter(avg_abv == max(avg_abv))


## -----------------------------------------------------------------------------
library(nycflights13)

# Information about flights
glimpse(flights)
# Information about airplanes
glimpse(planes)


## -----------------------------------------------------------------------------
# How many flights? How many planes?
c(nrow(flights), nrow(planes))

# How many flights have matching plane?
inner_join(flights, planes, by = "tailnum") %>% 
  nrow


## -----------------------------------------------------------------------------
# With left_join, we keep all flights
left_join(flights, planes, by = "tailnum") %>% 
  glimpse


## -----------------------------------------------------------------------------
# First we combine the two datasets 
# and create tot_delay variable
# Note: Could also use inner_join
dataset <- left_join(flights,        
                     planes,
                     by = "tailnum") %>% 
  mutate(tot_delay = dep_delay + arr_delay)


## ----error = TRUE-------------------------------------------------------------
# Next group by year and summarise
data_avg <- dataset %>% 
  group_by(year) %>% 
  summarise(avg_delay = mean(tot_delay, na.rm = TRUE))


## -----------------------------------------------------------------------------
# What happened?
# Both flights and planes have a variable year
# year.y refers to the one from planes
names(dataset)


## -----------------------------------------------------------------------------
# Try again
data_avg <- dataset %>% 
  group_by(year.y) %>% 
  summarise(avg_delay = mean(tot_delay, na.rm = TRUE))


## ----message = FALSE----------------------------------------------------------
library(tidyverse)

df_beers <- read_csv("beers.csv")
df_breweries <- read_csv("breweries.csv")


## -----------------------------------------------------------------------------
# Top 5 states for # breweries
state_top5 <- df_breweries %>% 
  count(state) %>% 
  top_n(5)


## -----------------------------------------------------------------------------
state_top5


## -----------------------------------------------------------------------------
breweries_top5 <- semi_join(df_breweries,
                            state_top5)


## -----------------------------------------------------------------------------
breweries_top5


## -----------------------------------------------------------------------------
# Only keep beers from these states
semi_join(df_beers,
          breweries_top5,
          by = "brewery_id") %>% 
  count(style, sort = TRUE)


## -----------------------------------------------------------------------------
# Let's look at the other states
breweries_nottop5 <- anti_join(df_breweries,
                               state_top5)


## -----------------------------------------------------------------------------
breweries_nottop5


## -----------------------------------------------------------------------------
# Only keep beers from these states
semi_join(df_beers,
          breweries_nottop5,
          by = "brewery_id") %>% 
  count(style, sort = TRUE)


## -----------------------------------------------------------------------------
library(nycflights13)

planes100 <- flights %>% 
  count(tailnum) %>% 
  filter(n >= 100)


## -----------------------------------------------------------------------------
flights100 <- semi_join(flights,
                        planes100)


## -----------------------------------------------------------------------------
# Do we get flights with missing
# tail number?
flights100 %>% 
  filter(is.na(tailnum)) %>% 
  nrow


## -----------------------------------------------------------------------------
# We can remove these NAs from planes100
planes100 <- filter(planes100,
                    !is.na(tailnum))
# Or we can remove them from flights100
flights100 <- filter(flights100,
                     !is.na(tailnum))


## ----eval = FALSE-------------------------------------------------------------
## inner_join(x, y, by = c("var1", "var2"))


## ----eval = FALSE-------------------------------------------------------------
## inner_join(x, y, by = c("name1" = "name2"))


## ----message = FALSE----------------------------------------------------------
library(tidyverse)
df1 <- tibble(
  x = c(1, 2),
  y = c(1, 1)
)
df2 <- tibble(
  x = c(1, 1),
  y = c(1, 2)
)


## -----------------------------------------------------------------------------
# Note that we get 3 rows, not 4
# because of duplicates
union(df1, df2)


## -----------------------------------------------------------------------------
intersect(df1, df2)


## -----------------------------------------------------------------------------
setdiff(df1, df2)


## -----------------------------------------------------------------------------
# The order is important!
setdiff(df2, df1)