dplyrGroupAndSort.r

# libraries and data
library(tidyverse)
library(dslabs)
data(murders)

# Demonstrate the use of groupby
# compute separate average and standard deviation for male/female heights
heights %>%
    group_by(sex) %>%
    summarize(average = mean(height), standard_deviation = sd(height))

# compute median murder rate in 4 regions of country
murders <- murders %>%
    mutate(murder_rate = total/population * 100000)
murders %>%
    group_by(region) %>%
    summarize(median_rate = median(murder_rate))

# calculate and extract the murder rate with one pipe
us_murder_rate <- murders %>%
    summarize(rate = sum(total) / sum(population * 100000) %>%
    .$rate

#Demonstrate the sorting of data
# set up murders object
murders <- murders %>%
    mutate(murder_rate = total/population * 100000)

# arrange by population column, smallest to largest
murders %>% arrange(population) %>% head()

# arrange by murder rate, smallest to largest
murders %>% arrange(murder_rate) %>% head()

# arrange by murder rate in descending order
murders %>% arrange(desc(murder_rate)) %>% head()

# arrange by region alphabetically, then by murder rate within each region
murders %>% arrange(region, murder_rate) %>% head()

# show the top 10 states with highest murder rate, not ordered by rate
murders %>% top_n(10, murder_rate)

# show the top 10 states with highest murder rate, ordered by rate
murders %>% arrange(desc(murder_rate)) %>% top_n(10)