Visualization in R

Base-R graphs

library(dplyr)

height_weight <- read.csv("/home/divenyijanos/Dropbox/teaching/Programming_Tools/Fall2015/Data/height_weight.csv")

## Warning in file(file, "rt"): cannot open file '/home/divenyijanos/Dropbox/
## teaching/Programming_Tools/Fall2015/Data/height_weight.csv': No such file
## or directory

## Error in file(file, "rt"): cannot open the connection

height_weight <- filter(height_weight, 100 < height, height < 250)

## Error in filter(height_weight, 100 < height, height < 250): object 'height_weight' not found

plot(height_weight)

## Error in plot(height_weight): object 'height_weight' not found

plot(height_weight$height, height_weight$weight)

## Error in plot(height_weight$height, height_weight$weight): object 'height_weight' not found

Package ggplot2

ggplot2 is a package developed by Hadley Wickham (so part of the “hadleyverse”). It is powerful visualization tool for R which is built upon the logic of grammar of graphics.

Most of the basic plots could be created by the “quick-plot” command qplot().

# install.packages("ggplot2")
library(ggplot2)
qplot(height, weight, data = height_weight)

## Error in ggplot(data, aesthetics, environment = env): object 'height_weight' not found

qplot(height, data = height_weight)

## Error in eval(aesthetics$x, data, env): object 'height_weight' not found

Quick-plot could be customized easily.

qplot(height, weight, data = height_weight, color = "darkred", size = 3)

## Error in ggplot(data, aesthetics, environment = env): object 'height_weight' not found

qplot(height, weight, data = height_weight, color = "darkred", size = 3, facets = ~ male)

## Error in ggplot(data, aesthetics, environment = env): object 'height_weight' not found

For complicated graphics, it is better to use the standard ggplot() formulation, which makes the underlying logic more explicit. What is behind a plot? We map different collections of data to different characteristics of a plot and use different visual elements to show them. In the graph above, we map the data about heigths to the x axis of the plot, the weights to the y axis (aesthetics), and use points to make them appear (geometries).

g <- ggplot(data = height_weight, aes(x = height, y = weight)) + geom_point()

## Error in ggplot(data = height_weight, aes(x = height, y = weight)): object 'height_weight' not found

## Error in eval(expr, envir, enclos): object 'g' not found

We can customize the geometries by setting the options.

g <- ggplot(data = height_weight, aes(x = height, y = weight)) + 
    geom_point(size = 3, color = "darkred")

## Error in ggplot(data = height_weight, aes(x = height, y = weight)): object 'height_weight' not found

## Error in eval(expr, envir, enclos): object 'g' not found

We can modify the general properties of the graph.

g + xlab("Height (cm)") + ylab("Weight (kg)") + ggtitle("Height vs weight")

## Error in eval(expr, envir, enclos): object 'g' not found

This formulation is great as it is really flexible. What if we would like to use different colors for males and females. This essentially requires the mapping of the male variable to the color of the points.

ggplot(data = height_weight, aes(x = height, y = weight, color = male)) + 
    geom_point(size = 3)

## Error in ggplot(data = height_weight, aes(x = height, y = weight, color = male)): object 'height_weight' not found

Here, factors help a lot. They show that our variable is categorical and could take only two values.

ggplot(data = height_weight, aes(x = height, y = weight, color = factor(male))) + 
    geom_point(size = 3)

## Error in ggplot(data = height_weight, aes(x = height, y = weight, color = factor(male))): object 'height_weight' not found

What if we would like to add a fitted line to the points? We could just add a new layer to the plot.

g + geom_smooth()

## Error in eval(expr, envir, enclos): object 'g' not found

g + geom_smooth(method = "lm")

## Error in eval(expr, envir, enclos): object 'g' not found

By carefully controlling the aesthetics of different geometries, we can fit the line either to each points or separately by gender.

ggplot(data = height_weight, aes(x = height, y = weight, color = factor(male))) + 
    geom_point(size = 3) +
    geom_smooth(method = "lm")

## Error in ggplot(data = height_weight, aes(x = height, y = weight, color = factor(male))): object 'height_weight' not found

ggplot(data = height_weight, aes(x = height, y = weight)) + 
    geom_point(aes(color = factor(male)), size = 3) +
    geom_smooth(method = "lm")

## Error in ggplot(data = height_weight, aes(x = height, y = weight)): object 'height_weight' not found

Here is a really good tutorial of the ggplot2 package and the grammar of graphics. It is also worth looking at the documentation which summarises all of the available options with lots of examples.