Instructions
- Homework assignment is here with all questions commented out.
- Complete all of the coding tasks in the
.qmd file
- Upload your individual exercise to your
GitHub repo by Wed 11:59pm.
- Remember to clean your github repo and sort hw submissions by weeks. Each week should have one folder.
# ##-------------------------------------------------------------------------
# ## R code for POLI502 Lab, week 5: Individual Exercises
# ## Written by: Howard Liu
# ##-------------------------------------------------------------------------
#
#
# # 1. Exploring a data set -------------------------------------------------
#
# # We have learned several functions to explore a data set, including
#
# dim(world.data)
# head(world.data)
# tail(world.data)
#
# # There are some other functions we can use. For example, the names
# # function tells us the names of all the variables included in a data frame
# # object.
#
# names(world.data)
#
# # The colnames function gives us the same results as well.
#
# colnames(world.data)
#
#
# # We can also apply the summary function without specifying variable
# # names. Then, R will provide the summary of ALL the variables in cluded
# # in a data frame object.
#
# summary(world.data)
#
#
# # The str() function tells us the structure of a data frame object,
# # meaning that it tells us which variables are factor, which ones are
# # numerical, which ones are logical, etc.
#
# str(world.data)
#
#
# # 2. Summarizing categorical variables ------------------------------------
#
#
# # The output from the str function above tells us that there are many factor
# # variables in the data set. For example, the democ_regime variable
# # is a factor variable (nominal-level). Summarize the information contained
# # in this variable by creating a frequency table.
#
#
# WRITE YOUR COMMAND HERE
#
#
# # The typerel variable is another factor variable.
# # This variable measures predominant religion in a given country.
# # Create a frequency table for this variable.
#
# WRITE YOUR COMMAND HERE
#
#
# # Make this frequency table vertical using the data.frame function
#
# WRITE YOUR COMMAND HERE
#
#
# # We have seen in the lecture that we often report RELATIVE frequencies
# # as well as raw frequencies. Relative frequencies can be obtained by
# # dividing each of the raw frequency values by the total number of
# # observations. Let's see how we do this.
#
#
# # To do so, it is better if we create a new object that stores the
# # frequency table. Let's create an object called ft.colony that is
# # equal to the vertical frequency table for the colony variable,
# # as follows.
#
# ft.colony <- data.frame( table(world.data $ colony) )
#
# # To make sure we did this correctly, let's take a look
#
# ft.colony
#
#
# # We can see that the first column, Var1, records all possible
# # values and the second column, Freq, records the raw frequency.
#
#
# # To convert the raw frequencies into relative frequencies, we divide
# # the values by the sum of Freq.
# # As we learned before, we use the sum function to calculate the sum of
# # all the values, as follows.
#
# sum( ft.colony $ Freq )
#
# # The relative frequencies are Freq divided by sum( ft.colony $ Freq )
#
# ft.colony $ Freq / sum( ft.colony $ Freq )
#
# # Alternatively, we can use the prop.table function to obtain the
# # same results
#
# prop.table(ft.colony $ Freq)
#
# # We would want to convert these further into percentages.
# # To make a ratio into a percentage, we simply multiply it by 100
#
# prop.table(ft.colony $ Freq) * 100
#
#
# # We would want to round these numbers to simplify the representation.
# # As we leanred two weeks ago, we use the round function to do that.
#
# round(prop.table(ft.colony $ Freq) * 100, digits = 2)
#
#
# # Finally, we want to insert these numbers into the frequency table
# # we created and stored in ft.colony.
#
# ft.colony
#
# # How do we do it?
# # We do this by creating a new column in the ft.colony object.
# # As we learned last week, we use the $ symbol to create a new column
# # in a data frame object, as follows
#
# ft.colony $ Percent <- round(prop.table(ft.colony $ Freq) * 100, digits = 2)
#
#
# # Now, our frequency table contains three columns, as follows
#
# ft.colony
#
# # Finally, we may want to change the column name for the first column
# # from "Var1" to something more intuitive.
# # To do so, we use the colnames function, as follows
#
# colnames(ft.colony)[colnames(ft.colony) == "Var1"] <- "Colonizer"
#
# ft.colony
#
#
# # We can see that about 33% of the countries in the world are former colonies
# # of the UK, about 15% of them are former colonies of France, about 10% of them
# # were never colonized, etc.
#
#
# # To summarize the steps to create a frequency table:
#
# ft.colony <- data.frame( table(world.data $ colony) )
# ft.colony $ Percent <- round(prop.table(ft.colony $ Freq) * 100, digits = 2)
# colnames(ft.colony)[colnames(ft.colony) == "Var1"] <- "Colonizer"
# ft.colony
#
#
#
# # Create a frequency table for the typerel variable
#
# WRITE YOUR COMMAND HERE
#
# # Which religion is the most "popular" in the world?
# # That is, what is the mode of the "typerel" variable?
#
# WRITE YOUR ANSWER IN WORD HERE
#
#
# # What is the percentage of countries where muslim is the majority?
#
# WRITE YOUR ANSWER HERE
#
#
# # Create a frequency table for democ_regime
#
# WRITE YOUR COMMAND HERE
#
# # What percentage of countries have a democratic regime?
#
# WRITE YOUR ANSWER IN WORD HERE
#
#
# # Create a bar chart to summarize the typerel variable
#
# WRITE YOUR COMMAND HERE
#
#
# # Create a bar chart to summarize the democ_regime variable
#
# WRITE YOUR COMMAND HERE
#
#
#
#
#
# # 3. Making ggplot graphs look nicer --------------------------------------
#
# # We have seen how to create a graph using the ggplot function
#
# ggplot(world.data, aes(x = colony)) + geom_bar()
#
# # The command above is the easiest way to produce a simple ggplot
# # graph, but we would want to modify some parts of the graph, such as
# # axis labels. For example, the graph above currently says
# # "colony" on the x-axis and "count" on the y-axis. We may want to
# # modify them so they can be more informative.
#
#
# # When we want to modify graphs, we usually create a ggplot graph
# # and store it into an object. Then we gradually add some features
# # to modify them. The above command can be re-written as follows:
#
#
# g <- ggplot(world.data) # Tells R which data frame contains the variable to plot
# g <- g + aes(x = colony) # Tells R which variable to plot
# g <- g + geom_bar() # Tells R what type of graph we want
# g # Tells R to show the contents of the object g
#
# # Now that we stored the graph into an object called g, we can
# # modify graph appearances by adding more options.
#
# # To change the label for the x-axis, we use the xlab option, as follows
#
# g <- g + xlab("Colony of Which Country?")
# g
#
# # Similarly, we can modify the label for the y-axis
#
# g <- g + ylab("Number of countries")
# g
#
# # If you want to change the text size for axes, do
#
# g <- g + theme(axis.text.x = element_text(size = 12))
# g
#
#
# # We can save this graph as a PDF file using the ggsave function.
#
# ggsave(file = "colony_bar.pdf", width = 10, height = 8)
#
# # The file option specifies the file name of the PDF file
# # you want to create.
# # The width and height option control the width and height of the
# # PDF file, respectively.
# # Once you save a graph in a PDF, you can easily embed it in a
# # Word document simply by drag & drop.
#
#
# # To summarize what we have done so far,
#
# g <- ggplot(world.data)
# g <- g + aes(x = colony)
# g <- g + geom_bar()
# g <- g + xlab("Colony of Which Country?")
# g <- g + ylab("Number of countries")
# g <- g + theme(axis.text.x = element_text(size = 12))
# g
# ggsave(file = "colony_bar.pdf", width = 10, height = 8)
#
#
#
#
# # Create a bar chart for the typerel variable, and save it as a PDF file
#
# WRITE YOUR COMMANDS HERE
#
#
#
#
#
#
# # 4. Summarizing numerical variables --------------------------------------
#
# # There are two variables in the data set, gini04 and gini08, that measure
# # the levels of economic inequality in a country numerically.
# # These are what's called Gini coefficient (Gini index or Gini ratio), which
# # takes values between 0 and 1 (or 0% and 100%). A value of 0 corresponds to
# # the "perfect equality" case, where everyone in a country is earning the same
# # amount of money, whereas a value of 1 (100%) corresponds to the maximal
# # inequality case, where one person is earning ALL the money in a country and
# # everyone else is earning nothing. The gini04 variable is from the year 2004
# # whereas the gini08 variable is from the year 2008.
#
#
# # Numerically summarize the gini04 variable.
# # That is, calculate and present the measures for central tendency and
# # those for dispersion.
#
# WRITE YOUR COMMANDS HERE
#
#
# # Numerically summarize the gini08 variable.
# # That is, calculate and present the measures for central tendency and
# # those for dispersion.
#
# WRITE YOUR COMMANDS HERE
#
#
# # Compare the distributions of gini04 and gini08. Do you think that the
# # level of economic inequality is getting worse, getting better, or neither?
# # Why or why not?
#
# # Note: I'm not asking why it is getting better, worse, or neither;
# # I'm asking on what basis you can conclude that it's getting better or worse.
#
#
#
# WRITE YOUR ANSWER HERE
#
#
#
# # Create a histogram of gini04
# # Modify the axis labels accordingly to make them informative and intuitive.
# # Save the graph as a PDF file.
#
# WRITE YOUR COMMANDS HERE
#
#
#
# # Create a histogram of gini08
# # Modify the axis labels accordingly to make them informative and intuitive.
# # Save the graph as a PDF file.
#
# WRITE YOUR COMMANDS HERE
#
#
#
# # Compare the distributions of gini04 and gini08 graphically by
# # placing the two PDF files you just created side by side.
# # Do you confirm the conclusion you derived previously?
#
#
# WRITE YOUR ANSWER HERE
#
#
#
#
# # As we saw in the lecture, we sometimes create histograms for different
# # values of a nominal-level variable. For example, we may want to create
# # separate histograms of gini04 for countries in different regions.
#
#
# # To do so, we use the facet_wrap option, as follows.
#
#
# COPY & PASTE THE CODE YOU WROTE TO PRODUCE HISTOGRAM FOR gini04 HERE
#
# g <- g + facet_wrap( ~ region)
# g
#
#
# # We can see that Scandinavian countries and Western European countries
# # have, on average, lower Gini coefficients (= they are more egalitarian),
# # whereas countries in South America have relatively high values.
#
#
#
#
# # Create separate histograms of women09 for countries in different regions.
#
#
# WRITE YOUR ANSWER HERE
#
#
#
#
# # We may want to do the same using numerical methods.
# # That is, we may want to obtain central tendencies and dispersions for
# # a numerical variable for different groups.
#
# # To do so, we use the by function.
# # The by function take the following form
# #
# # by( VARIABLE_YOU_WANT_TO_ANALYZE, GROUP, FUNCTION )
# #
# # That is, you provide
# # (1) an interval-level variable you want to summarize first,
# # (2) a comma
# # (3) a nominal variable that separates observations into groups
# # (4) a comma
# # (5) a function you want to apply (such as summary, mean, median, sd, etc.)
#
# # For example, to obtain numerical summaries of gini04 for different regions,
# # we write
#
# by(world.data $ gini04, world.data $ region, summary)
#
#
#
# # Calculate the standard deviation of gini04 for different regions using
# # the by function
# # Hint: we still need to take care of the missing value problem.
# # Use the na.rm = TRUE option.
#
#
# WRITE YOUR COMMAND HERE
#
#
# # Which region has the smallest dispersion?
#
#
# WRITE YOUR ANSWER HERE
#
#
# # End of file