Week 7: Apply it to your data 6

Introduction

Questions

Variation

Visualizing distributions

ggplot(data = myData) +
    geom_bar(mapping = aes(x = country))

myData %>% count(country)

## # A tibble: 7 × 2
##   country     n
##   <chr>   <int>
## 1 CHINA      45
## 2 FRANCE    210
## 3 INDIA       3
## 4 PAKIST      2
## 5 UK         45
## 6 USA      1032
## 7 USSR      714

ggplot(data = myData) +
    geom_histogram(mapping = aes(x = latitude))

## `stat_bin()` using `bins = 30`. Pick better value `binwidth`.

ggplot(data = myData, mapping = aes(x = latitude, colour = country)) +
    geom_freqpoly()

## `stat_bin()` using `bins = 30`. Pick better value `binwidth`.

Typical values

myData %>%
    filter(!is.na(latitude)) %>%
    ggplot(aes(x = latitude)) +
    geom_histogram(binwidth = 0.1)

# Wider bin to compare
myData %>%
    ggplot(aes(latitude)) +
    geom_histogram(binwidth = 0.5)

Unusual values

myData %>%
    ggplot(aes(x = latitude)) +
    geom_histogram()

## `stat_bin()` using `bins = 30`. Pick better value `binwidth`.

myData %>%
    ggplot(aes(x = latitude)) +
    geom_histogram() +
    coord_cartesian(ylim = c(0, 50))

## `stat_bin()` using `bins = 30`. Pick better value `binwidth`.

Missing values

myData %>%
    # Only impossible latitudes become NA (valid range is -90 to 90)
    mutate(latitude = ifelse(latitude < -90 | latitude > 90, NA, latitude)) %>%
    ggplot(aes(x = longitude, y = latitude)) +
    geom_point()

Covariation

A categorical and continuous variable

myData %>%
    ggplot(aes(x = country, y = latitude)) +
    geom_boxplot() +
    coord_flip()