Hello! Now we are going to analyse some data using plots

In this post we have done some wrangling to form groups and analyse them in a numerical way. Today we are going to create some plots that helps us with the interpretation of data

We will remain using the data from mpg data set included in the ggplot package, so let’s load the tidyverse:

library(tidyverse)

To a better understandment of the data, click on the link to read about it in the previous post.

“How motors improved during these years?”

To answer this question, we can plot the efficiency of the motors (measured by city miles per gallons), let’s use ggplot2 to do that.

Histograms

Histograms are also known as frequency plots, and they show us how many times a value appears in our data set:

mpg %>%
  ggplot(aes(x = cty)) +
  geom_histogram(binwidth = 1)

Notice that in the Y axis is the count of how many times that value appears, while in the X axis we can check the values of cty (city miles per gallons)

We can built more histogram and also compare them:

mpg %>%
  filter(year == 1999) %>%
  ggplot(aes(x = cty)) +
  geom_histogram()

mpg %>%
  filter(year == 2008) %>%
  ggplot(aes(x = cty)) +
  geom_histogram()

Now we can use the Patchwork package to combine the visualisation of both plots:

library(patchwork)

hist1999 <- mpg %>%
  filter(year == 1999) %>%
  ggplot(aes(x = cty)) +
  geom_histogram() +
  xlab("1999")
  
hist2008 <- mpg %>%
  filter(year == 2008) %>%
  ggplot(aes(x = cty)) +
  geom_histogram() +
  xlab("2008")
  
hist1999  + hist2008

Interesting, right?

We can also change the colour of our plots to improve comparison:


hist1999 <- mpg %>%
  filter(year == 1999) %>%
  ggplot(aes(x = cty)) +
  geom_histogram(color = "Blue", binwidth = 1) +
  xlab("1999")
  
hist2008 <- mpg %>%
  filter(year == 2008) %>%
  ggplot(aes(x = cty)) +
  geom_histogram(color = "Red", binwidth = 1) +
  xlab("2008")
  
hist1999  + hist2008

Watching it like this makes it easier to conclude that cars have become more efficient during years.

There are other interesting ways of improving the quality of our histograms, check how they come out with colours:

colour_hist1999 <- mpg %>%
  filter(year == 1999) %>%
  ggplot(aes(x = cty)) +
  geom_histogram(colour = "Firebrick", binwidth = 1, fill = "MidnightBlue") +
  xlab("1999")
  
colour_hist2008 <- mpg %>%
  filter(year == 2008) %>%
  ggplot(aes(x = cty)) +
  geom_histogram(colour = "dodgerblue", binwidth = 1, fill = "Firebrick") +
  xlab("2008")
  
color_hist1999  + color_hist2008

A similar type of plot that also has plenty of information about groups is the density plot.

Density plots

With ggplot we can plot density of out data in a simple way:

library(tidyverse)
mpg %>%
  ggplot(aes(x = cty)) +
  geom_density(aes(colour = as.factor(year))) +
  xlab("Density Plot")

In this kind of plot we can see the presence of many subgroups, the area under the curves of each subgroups sums to 1. This allows us to compare subgroups of diferent sizes.

It’s also possible to optimise visualisation of this plot, check the argumentsof the funtion geom_density() and explore what it is capable of doing!

Plotting with ggplot2

Hello! Now we are going to analyse some data using plots

“How motors improved during these years?”

Histograms

Density plots