Our goal in this first set of exercises is to produce some plots and learn a bit about how R/RStudio works.
Make sure you have both R and RStudio installed. In addition, make
sure you have the tidyverse package installed.
Both of these are detailed here:
Start by downloading labA01.R to somewhere on your computer. The easiest way to do this is to right click on the URL below and “Save file as…”
https://www.massey.ac.nz/~jcmarsha/161122/labs/labA01.R
Then load it into RStudio. The fail-safe way to do that is to first load RStudio up, and then use the File->Open File… option (alternatively the folder button on the toolbar), browsing to where the file was saved, and loading it up.
Alternatively, you may be able to right-click on the file in explorer or Finder, then Open with… and choose RStudio.
When done, you’ll see that RStudio loads the script (a text file with
a list of comments and commands) into the Editor pane in
the top right. There’s 3 other panes available: The Console
pane in bottom left, which is where commands are run and (text) output
given, The Environment/History pane in top right, and the
Files/Plots/Packages/Help pane in bottom right. We’ll use
all of them today!
Start by reading through the script. You’ll see it has a bunch of comments at the top (each line beginning with a hash character) which explain what is going on, followed by some lines of code that perform various tasks.
The first command is library(tidyverse). This tells
RStudio to load in the libraries (also known as packages) called the
tidyverse. Packages are a way to extend the base R/RStudio
functionality. In this case, the tidyverse package is a set of packages
for doing data analysis in a consistent way. See http://www.tidyverse.org
for more information.
Place your cursor on the line library(tidyverse) and
then click on the Run button in the top right of the editor
pane. Alternatively, Ctrl-Enter
(Command-Enter on a Mac) will run the line.
What should happen is the line will be copied down to the
Console window, where it will be executed. In this case,
the package will be loaded and you should get a message that looks
something like this:
## ── Attaching packages ─────────────────────────────────────── tidyverse 1.3.1 ──## ✔ ggplot2 3.3.5 ✔ purrr 0.3.4
## ✔ tibble 3.1.6 ✔ dplyr 1.0.8
## ✔ tidyr 1.2.0 ✔ stringr 1.4.0
## ✔ readr 1.4.0 ✔ forcats 0.5.1## ── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
## ✖ dplyr::filter() masks stats::filter()
## ✖ dplyr::lag() masks stats::lag()If you get an error here, it’s probably due to the
tidyverse packages not being installed - check https://r-resources.massey.ac.nz/help/usingrin161.122.html
or contact your lecturers to get things figured out.
Now, we’re going to load up some data on earthquakes from Fiji.
This dataset is included with R/RStudio. Place your cursor on the
data(quakes) line and run it. Again, the code will be
copied down to the Console and executed. You’ll see you now have a
quakes object in the Environment pane in the
top left.
Double click on the quakes object in the
Environment pane to have a look - this shows you the data
in a spreadsheet form. You should see 5 columns of data with
observations from 1000 earthquakes.
The next 2 lines of code produce a plot. This seems quite a lot of effort! But that effort will be rewarded by allowing us to produce very elegant charts that tell clear stories later on!
To run the two lines, you can place your cursor anywhere within those
2 lines and hit the Run button or
Ctrl-Enter
(Command-Enter on Mac). You can also highlight
the 2 lines and do the same thing if you prefer. Notice that R/RStudio
is treating the 2 lines as if they were a single statement - in fact,
they are! The two parts are joined up using the + operator
in this case. Essentially we “add” the geom_point bit on to
the ggplot bit to produce the plot.
Let’s break down what is going on, as it’s quite a lot! One way to do that is to run it in sections.
Type in (or copy + paste) ggplot(data=quakes) on a
separate line in the R script and run that line.
This tells R/RStudio to create a new plot using quakes
as the data - it’ll just give a blank canvas in the plot window.
NOTE If you accidentally added the +
from the original, when you run it you’ll find that it copies to the
Console but you’ll be left with a + prompt
there, rather than the usual > prompt. This is R/RStudio
knowing that the command is not complete, so it’s waiting for the rest
of it. If this happens and you don’t intend it, just hit
Esc key until the prompt returns to the usual
>.
Next, type in the bit inside the geom_point function:
aes(x=long, y=lat) on a separate line in the R script and
run that. You should get:
## Aesthetic mapping:
## * `x` -> `long`
## * `y` -> `lat`defining the mapping from “Aesthetic” features of the plot (e.g. the
x and y axes) to features of the data. The aes is short for
aesthetics.
Lastly, the geom_point() command adds a layer of
geometry to the plot, mapping from the data to the axes using the
aes bit. In this case we’re adding points. So this will add
each observation as a point on the plot.
In addition to the x and y aesthetics we’ve
used above, there’s a bunch of others. For example, you can change
colour, shape, size as well as
transparency of points using alpha. Plus lots more.
Try the following:
ggplot(data=quakes) +
geom_point(mapping=aes(x=long, y=lat), col='pink', size=2)Notice the 'pink' there is in quotation marks. This is
because it’s a label for something, not a quantity that R/RStudio would
already know about. You can use either single quotes ' or
double quotes " as long as you’re consistent (i.e. start
and end with the same thing).
Also, notice that when we specify multiple parameters to the function
geom_point() we separate them using a comma. This is a
general principle: functions in R/RStudio always use parentheses when we
run them, and parameters that we specify are always separated by commas.
Note we did that with the aes() function too.
You might want to experiment with the other parameters. What
happens when you use shape=2 for example? Or
alpha=0.4? Try experimenting!
In the same way that x and y are mapped
to features in our data, we can do the same for the other aesthetics.
Try adding col=mag to the aes() function so
that it looks like:
ggplot(data=quakes) +
geom_point(aes(x=long, y=lat, col=mag))You might want to try looking at col=stations as
well. You should find they look really similar. This makes sense -
larger earthquakes are likely felt more widely, so will be detected at
more stations.
Once you’re done, remember you can add comments to your script - save it in a safe place so you can go back to it later.