--- title: "Lab-2: Data Wrangling II + Relationships I + II" --- # Instructions - You will do your work in this `Quarto` document but will submit all your answers on Canvas - *very important*: As you move through the document, make sure to run all code chunks (grey bits) that already have code in them - Write your own code in the empty code chunks - Useful shortcuts: - to run all the code in a specific code chunk, press the green right-facing triangle at the top right of the code chunk - to run all *prior* code chunks, press the downward-facing gray triangle at the top right of the code chunk ```{r load-libraries, warning=FALSE, message=FALSE} library(tidyverse) library(socviz) library(juanr) ``` # Task 1: Pokemon First we'll look at data on pokemon. Let's load it below and look at it. ```{r load-pokemon} pokemon ?pokemon ``` Finally, let's create some new variables and do some plotting. Construct a variable called `diff` that equals a Pokemon's attack level minus their defense level. This variable tells you how much better (or worse) a Pokemon is at attack than defense. ```{r} ``` Make a boxplot graph where you compare the distribution of `diff` across a Pokemon's `type1`. ```{r} ``` **Question 1: right click the graph --\> "save image as" --\> and save it somewhere on your computer. You will need to submit the plot!** **Question 2: which type has the MOST outliers in terms of Pokemon who are better at defense than attack?** # Task 2: Opiate deaths Let's look at data on deaths related to opiate use in the US, from the `socviz` package. ```{r} opiates ``` First, make a barplot of opiate deaths (x-axis) by state (y-axis) but only for one year, 2010. ```{r} ``` **Question 3: What's the top state in terms of opiate deaths?** The plot lets you compare the number of opiate deaths across states, but there's a problem: some states have a lot more people, and will therefore have *more* deaths even if the opiate crisis isn't necessarily worse there. To account for this, construct a new variable gives you the death *rate* per 100,000 residents. To get the rate: - divides deaths by population - multiplies that number by 100,000 - call it `deaths_per_100` This gives you the death *rate* per 100,000 residents. Make the same barplot as above but this time with your new *rate* variable on the y-axis: ```{r} ``` **Question 4: Which state has the worst opiate death rate?** **Question 5: Which of the following is closest to the meaning of this state having the highest opiate death rate:** - this state has the most deaths overall in 2010 - this state has the most deaths relative to it's population in 2010 - this state has the highest year-to-year increase in opiate deaths in 2010 - the rate of opiate deaths in this state has been increasing over time # Task 3: World Bank categories Let's look at data on countries' levels of national income per capita (GNI per cap): ```{r} income = read_rds("https://github.com/hail2thief/poli-301/raw/master/static/files/gni-2021.rds") %>% # drop missing income data drop_na(gni_cap) income ``` The World Bank likes to categorize countries into "buckets" based on national income. The buckets are as follows: - Low income = less than \$1,046 - Lower-middle = \$1,046 – 4,095 - Upper-middle = \$4,096 - 12,695 - High income = \$12,695 Use `case_when` to construct a new variable called `category` that categorizes countries into these buckets. Store your new data as an object named `income_cats`: ```{r} ``` How many countries are in the "Lower-middle" category? Run the code below. The code will tell you how many countries are in each `category`: ```{r, eval = FALSE} income_cats %>% group_by(category) %>% tally() ``` **Question 6: What percent of countries are in the low income category?** # Task 4: Money in politics We're going to work with the Database on Ideology, Money in Politics, and Elections (`bonica`), which has candidate-level information on campaign fundraising. ```{r} bonica ?bonica ``` How much money did the average candidate from each party raise in the 2018 election cycle? Calculate the average amount of money raised (`total_receipts`) per political party in the 2018 cycle. ```{r} ``` **Question 7: How much money did the TOP fundraising party raise, on average?** Look only at California politicians. Among these, who raised the most money? Hint: you may need to use two filters sequentially instead of one. ```{r} ``` **Question 8: Who raised the most money?** How much do the number of donors vary *within* states? For the 2018 cycle, calculate the standard deviation of the number of donors each candidate received money from (`num_distinct_donors`), broken down by state. Then, look at which state has the highest standard deviation of the number of distinct donors. ```{r} ``` **Question 9: Which state has the highest standard deviation of distinct donors?** The reason this state has such a high standard deviation of the number of unique donors is because of one candidate who is unusually good at fundraising. Dig around in the data for this state; who is this person? ```{r} ``` **Question 10: Who is this person?** **Question 11: What can we infer from the fact that this state has the highest standard deviation of the number of unique donors?** # Task 5: Leaders Look at the leader dataset on world leaders: ```{r} leader ?leader ``` Across the whole dataset, how many years of experience does the median leader have? ```{r} ``` **Question 12: how many years of experience does the median leader have?** What percent of world leaders have experience in a rebellion in the year of 1991? ```{r} ``` **Question 13: What percent of world leaders in 1991 had experience in a rebellion?** How is the age of world leaders changing over time? Calculate the average age of world leaders per year. Plot the average ages as a lined time series. ```{r} ``` **Question 14: right click the graph --\> "save image as" --\> and save it somewhere on your computer. You will need to submit the plot!** # Task 6: bots Look at the bot data on how Americans feel about different political issues: ```{r} bot ?bot ``` What percent of of respondents in `bot` support concealed carry? ```{r} ``` **Question 15:**  **What percent of WHITE respondents in `bot` support concealed carry?** Let's look at how the different religious groups feel about abortion (`prochoice`). Calculate the percent of respondents who support or oppose `prochoice`, broken down by religion. Store as an object so you can look at it. ```{r} ``` **Question 16: What religious group has the highest opposition to `prochoice`?** # All Done!