Please answer the questions below in the R markdown file (.Rmd) provided.  You must download the R markdown file and enter your answers to the questions under the relevant question sections.  Remember to create a new project folder with the correct folder structure!  Download the data required for this assignment and put it in your 'data' folder.  When you are happy with your answers, upload the R markdown file you created (the one that contains your answers to the questions) and any figures that were saved as part of the assignment.  Don't forget to hit the 'submit' button once you have uploaded everything.  You are allowed unlimited submission attempts before the deadline, so don't worry if you forgot to upload something or forgot to answer a question, you can make changes and upload a new submission any time before the deadline.  Make sure to save your work on your own PC (or to your OneDrive account) so that you don't lose your work.

Before answering the questions, test out the R markdown file to see if it will 'knit' to html output from the outset.  Once you have that working, you can make your additions/changes in order to answer the questions below.  It is good practice to test the html 'knit' each time you make a change to your R markdown file to make sure that your code is correct and you haven't made any errors or typos etc.  There may be some questions you cannot answer.  If this happens and a it means you can't test whether your R markdown file will 'knit' correctly for a subsequent question, use the comment function to comment out the R code that isn't working.  If a subsequent answer depends on having created a column from a previous question to which you did not know the answer (i.e., you couldn't work out how to create the new column), use an existing column from the tibble to answer the subsequent question.  Your answer will not receive a full score, but if you get the syntax correct and just insert the wrong column you will get most of the points for that question.

Please do your best to structure your code well with useful commenting and indentation, section headers, etc.

Question 1:

Adjust the 'author' and 'date'  of your R markdown file to reflect your own name and the date on which you started to work on the assignment. [2 points]

Does your .Rmd file 'knit' to html after you downloaded it?  If not, make the necessary changes/additions in order to make it work? (Hint: take a look at the portion of the markdown file where packages are being loaded and think about how that relates to the error message you receive). [2 points]

Question 2:

After loading the data as a tibble, carry out a random inspection of the contents of the tibble (the data).  Insert code under the section corresponding to Question 2 to do this 4 times, each time using a different function. [4 points]

While inspecting the data, did you notice anything about the data that you think is important to keep in mind during your data analysis?  If so, does this apply to all columns, or only some columns in the tibble?  Please explain your reasoning. [2 points]

Please describe any differences you observe between the output in the html file produced by each of the 4 functions.  Write your answers in the allocated space in the R markdown file. [5 points]

Question 3:

The data you downloaded are taken from the Subtlex-UK database.  Subtlex is a project designed to collect large corpora of statistics related to lexical items, e.g.,  ratings, scores, counts/frequencies, etc.  These are derived from millions (or these days billions) of instances of the lexical items in the database, originally taken from book or newspaper subtitles.  There are separate corpora for different languages, and even different dialects of the same language (hence the 'UK' part refers to data on UK English).  This corpus is actually based on data taken from BBC British television programs, which are considered to be more representative of the current language use in the UK.  You can read more about these data if you are interested in the following paper.  We are not going to work with all the data from this database, so let's make a sensible sub-selection using the 'dplyr' functions we learned about in class.

Important: In order to get a full score your answer should use the '%>%' to create a pipeline for the different data wrangling steps in a-e below.

First, select the columns with the following names (make sure the new/updated tibble has the columns in the order in which they are mentioned here): 'Spelling', 'DomPoS', 'BNC_freq', 'Zipf'. [2 points]

Next, filter the data based on the 'DomPoS' column to include only adjectives and nouns in the tibble. [2 points]

Create a new column in the tibble called 'FreqLog10' that takes the base 10 logarithm of the frequency of each lexical item (Hint: the frequency is in the 'BNC_freq' column). [2 points]

Rename the columns to have more intuitive and convenient column labels.  Call the first column 'Word'; the second column 'PoS' (this indicates Part of Speech); the third column 'Freq' (this indicates word frequency). [2 points] 

Sort the tibble in reverse alphabetical order according to the 'Word' column. [2 points] 

Do you notice anything strange about the 'FreqLog10' column?  Try to work out why by comparing it to the 'Freq' column and share your conclusion. [2 points]

Question 4:

Now create some figures using ggplot2 to explore the data and at each step display the figure you created in your html output file.  You must use ggplot2 to get a full score.

Create a figure called 'fig1' that uses geom_text() to display the word frequency ('Freq' column in our tibble) on the x axis and the corresponding log transformed lemma frequency ('Zipf' column in our tibble) on the y axis.  For the text labels use the 'PoS' column and set the theme for the plot to theme_minimal(). [5 points]

Create a new figure called 'fig2' that uses geom_boxplot() to display the frequency ('Freq' column in our tibble) on the y axis as a function of part of speech ('PoS' column in our tibble) on the x axis.  Before plotting first exclude frequency values higher than 1000.  Set the theme for the figure to theme_light(). [6 points]

Use the grid.arrange() function to combine 'fig1' and 'fig2' into a single figure (fig1 above fig2) and call it 'fig3'.  Display 'fig3' in your html output as well. (Hint: don't forget to load the 'gridExtra' library). [3 points]

Use the ggsave() function to save 'fig3' into the 'figures' folder in your project folder with the filename 'fig3.png'. (Hint: Use ?ggsave() to inspect the R help documentation for this function to see how to specify the argument for selecting the correct figure). [2 points]

Question 5:

Go through your R markdown document and change the section headings "Question 2", "Question 3", and "Question 4" from minor titles to major titles in the html output. [1 point]

Go to line12 in your R markdown file and make the text there display as bold text after you 'knit' the file to html. [1 point]