Week 7: Apply it to your data 6

Import Data

myData <- read_xlsx("../00_data/myData.xlsx")

## New names:
## • `` -> `...7`

Introduction

Questions

Variation

Visualizing distributions

myData %>%
    ggplot(aes(x = alcohol)) + 
    geom_bar()

myData %>%
    ggplot(mapping = aes(x = quality)) +
    geom_histogram(binwidth = .5)

myData %>%
    filter(quality > 4) %>%
    ggplot(aes(x = quality)) + 
    geom_histogram(binwidth = .05)

myData %>% 
    ggplot(mapping = aes(x = quality, color = alcohol)) + 
    geom_freqpoly()

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

Typical values

myData %>%
    
    # Filter out diamonds > 3 carat
    filter(quality > 4) %>%
    
    # Plot
    ggplot(aes(x = quality)) + 
    geom_histogram(binwidth = 1)

Unusual values

myData %>% 
    ggplot(aes(pH)) + 
    geom_histogram()

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

myData %>% 
    ggplot(aes(pH)) + 
    geom_histogram() + 
    coord_cartesian(ylim = c(0, 1))

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

Missing Values

Covariation

myData %>%
    ggplot(aes(x = alcohol, y = fixed_acidity)) + 
    geom_boxplot()

## Warning: Continuous x aesthetic -- did you forget aes(group=...)?

A categorical and continuous variable

myData %>%
    ggplot(aes(x = alcohol, y = fixed_acidity)) + 
    geom_boxplot()

## Warning: Continuous x aesthetic -- did you forget aes(group=...)?

Two categorical variables

myData %>%
    
    filter(alcohol < 14, alcohol > 5) %>%
    count(quality, alcohol)%>%
    ggplot(aes(x = quality, y = alcohol, fill = n)) +
    geom_tile()

Two continous variables

myData %>%
    ggplot(aes(x = quality, y = fixed_acidity)) + 
    geom_hex()

myData %>%
    filter(quality > 3) %>%
    ggplot(aes(x = quality, y = fixed_acidity)) +
    geom_boxplot(mapping = aes(group = cut_width(quality, .1)))

## Patterns and models

library(modelr)
mod <- lm(log(fixed_acidity) ~ log(quality), data = myData)

myData2 <- myData %>%
    modelr::add_residuals(mod) %>%
    mutate(resid = exp(resid))

myData2 %>%
    ggplot(aes(quality, resid)) +
    geom_point()

myData2 %>%
    ggplot(aes(alcohol, resid)) +
    geom_boxplot()

## Warning: Continuous x aesthetic -- did you forget aes(group=...)?