Log Week 2

Step 1 - The ‘here’ Package

First I must install, load and test the here package.

install.packages("here")

library(tidyverse)
library(here)
here::here()

## [1] "/Users/lukekeevers/Desktop/Uni/Third Year/T1/3361 (Internship)"

here("data", "chickflick.csv")

## [1] "/Users/lukekeevers/Desktop/Uni/Third Year/T1/3361 (Internship)/data/chickflick.csv"

Aha! I have realised that including the “install.packages” function makes it so the package is re-installed every time I knit the document. So I will include “eval=FALSE” into the code chunk to stop it from running this code after I have installed the package.

Honestly, I’m not really sure what the “here” function just did but it seems to load the data correctly…

Step 2 - Making a Bar Graph

chickflick <- read.csv(here("data", "chickflick.csv"), header = TRUE)

I will be using what I have learnt from Danielle’s guides to create a graph which visualises the different arousal levels of men and women to two movies. I got this example data from Dr Erin Buchanan’s site here.

First I will factor the variables appropriately.

chickflick$gender = factor(chickflick$gender,
                           levels = c(1,2),
                           labels = c("Men", "Women"))
chickflick$film = factor(chickflick$film,
                         levels = c(1,2),
                         labels = c("Bridget Jones", "Memento"))

Time to create the bar chart!

chickbar <- ggplot(data = chickflick, 
                   mapping = aes(x = film, y = arousal, fill = gender))
print(chickbar)

Okay…so I knew this wouldn’t produce the final product but this is less than I was expecting. I forgot about what Danielle said about plots in ggplot using “layers”. So I went online and figured out how to add my mean bars and some confidence interval error bars.

Step 3 - Actually Making It…

So the “stat_summary” function allows me to add my layers. The “geom” argument specifies what object I want to represent my selected function, “fill” changes the colour of my object, and “colour” changes the colour of the outline.

For the error bars, the “width” argument allows me to change the width of the bars, whilst the “position =”dodge"" part makes sure parts of the graph don’t get in the way of each other. Finally, I add some x and y labels and… now I get a much nicer looking graph.

chickbar <- chickbar + 
  stat_summary(fun.y = mean,
               geom = "bar",
               position = "dodge",
               colour = "black") +
  stat_summary(fun.data = mean_cl_normal,
               geom = "errorbar",
               width = .2, position = position_dodge(width = 0.90)) +
  xlab("Film Watched") +
  ylab("Arousal Level")

print(chickbar)

Step 4 - Final Improvements

It is time to make this nice looking graph a bit more science-journal-appropriate. First I can fix the legend to have more appropriate labels and I can change the colour of the bars. Even though I don’t see many colourful graphs in the papers I read, I want mine to actually be interesting to look at. I found a list of all the ggplot colours here.

chickbar <- chickbar +
  scale_fill_manual(name = "Gender of Participants",
                    labels = c("Male", "Female"),
                    values = c("deepskyblue1", "magenta1"))
print(chickbar)

I also wanted to know how to get rid of the ugly grey background. Luckily, the internet has a solution to that too! I think I will be using these particular lines of code often in the future, so I will assign them their own term - “background”.

background <- theme(panel.grid.major = element_blank(),
                    panel.grid.minor = element_blank(),
                    panel.background = element_blank(),
                    axis.line = element_line(colour = "black"))

chickbar <- chickbar + background

print(chickbar)

And finally - add a title!

chickbar <- chickbar + 
  ggtitle(label = ("Arousal of Men and Women to Two Hit Movies"),
          subtitle = ("A Nobel Prize-deserving finding"))

print(chickbar)