## Problemset Title ---- ## [ PROJ ] < Problem set 4 > ## [ FILE ] < lastname_firstname_problemset4.R > ## [ AUTH ] < Your name / GitHub handle > ## [ INIT ] < Date you started the file > rm(list=ls()) # clears environment so script runs from a clean slate ## Overview ---- # 26 points # In last week's problemset, you used `dplyr` functions such as `filter()`, # `arrange()`, and `select()` and the pipe operator `%>%` to perform data manipulations. In this # problemset, you'll practice creating variables using `mutate()` in combination with # `if_else()`, `case_when()`, and `recode_values()`. # NOTE ON BACKTICKS: In these instructions, function and variable names are wrapped in backticks (a common practice in R) # (e.g., `mutate()`, `df_schools`) just to show they are code. Do NOT type the backticks in your # own code. Write mutate(...) and df_schools, not `mutate()` and `df_schools`. # Backticks are not the same as quotes: "text in quotes" is a character value, which you'll still # use for things like state_code == "CA". ## Question 1: Data manipulation using pipes ---- # Q1.1 ---- # 1 point # First, load the `tidyverse` library. # Second, use the `load()` and `url()` functions to download the `df_school_all` dataframe from this URL: # https://github.com/anyone-can-cook/rclass1/raw/master/data/recruiting/recruit_school_allvars.RData # Notes on data: Each row in `df_school_all` represents a high school (includes both public and private) # There are columns (e.g., `visits_by_100751`) indicating the number of times a university visited that high school # The variable `total_visits` identifies the number of visits the high school received from # all (16) public research universities in this data collection sample # Q1.2 ---- # 1 point # Use the functions `arrange()`, `select()`, and `head()` to do the following: # !!!! Complete the tasks using pipes `%>%` !!!! # 1. Sort `df_school_all` descending by `med_inc` # 2. Select these variables, in this order: # `name`, `state_code`, `city`, `school_type`, `total_visits`, `med_inc`, `pct_white`, # `pct_black`, `pct_hispanic`, `pct_asian`, `pct_amerindian` # 3. Show the first 10 rows of the dataframe, which represents the schools that have students with the top 10 median income # 4. Assign this to a new object named `q1_2` q1_2 <- # Q1.3 ---- # 1 point # Building upon the previous question, use the functions `arrange()`, `select()`, `filter()`, # and `head()` to obtain observations on public high schools in New York with the top 10 # median income. Select the same variables, in the same order, as in Q1.2. # Assign this data to a new object named `q1_3` # HINT: Before filtering, check how the values are spelled, e.g. # df_school_all %>% count(school_type) # HINT: `filter()` must come before `head()`. Otherwise you are filtering the national top 10. q1_3 <- # Q1.4 ---- # 1 point # Now obtain the top 10 observations of private high schools in New York # according to highest median income. Select the same variables, in the same order, as in Q1.2. # Assign this data to a new object named `q1_4` q1_4 <- ## Question 2: Variable creation using `tidyverse`'s `mutate()` ---- # Often before creating new "analysis" variables, you may want to investigate the values of # "input" variables. Here are some examples of checking variable values using `count()`: df_school_all %>% count() df_school_all %>% filter(is.na(med_inc)) %>% count() df_school_all %>% count(school_type) ### The following questions will ask you to use `mutate()` with `if_else()` # to create a 0/1 indicator and then use `count()` to generate frequency tables. # HINT: the general pattern is # df %>% mutate(new_var = if_else(condition, 1, 0, missing = NULL)) # where `condition` is something that is TRUE or FALSE for each row (e.g., state_code == "XX"). # Use 1 and 0 (not TRUE/FALSE) for your indicator values. # Q2.1 ---- # 1 point # Create a 0/1 indicator called `ca_school` for whether the high school is in California. # Start from `df_school_all` and assign the full dataframe (all original columns plus your new # `ca_school` column) to a new object named `q2_1`. q2_1 <- # Q2.2 ---- # 1 point # Generate a frequency table with `count()` for the `ca_school` variable you just created. # Assign this frequency table to a new object named `q2_2` q2_2 <- # Q2.3 ---- # 1 point # Create a 0/1 indicator called `ca_pub_school` for whether the high school is # a public school in California. # Assign the full dataframe with your new column to a new object named `q2_3`. # HINT: both conditions go inside one `if_else()`, joined with `&`. q2_3 <- # Q2.4 ---- # 1 point # Generate a frequency table with `count()` for the `ca_pub_school` variable you just created. # Assign this frequency table to a new object named `q2_4` q2_4 <- ## Question 3 ---- # Complete the following steps to create an analysis variable using `mutate()` and `if_else()`: # Q3.1 ---- # 1 point # First, use `select()` to select `name`, `pct_black`, `pct_hispanic`, `pct_amerindian` from # `df_school_all`, and assign to a new object named `df_race`. df_race <- # Q3.2 ---- # 3 points # Starting from `df_race`, use `filter()`, `is.na()`, and `count()` to investigate whether or not # `pct_black`, `pct_hispanic`, `pct_amerindian` have missing values. # Assign each of these investigations to objects named `q3_2_a`, `q3_2_b`, and `q3_2_c`, respectively. # HINT: look at the second example at the start of Question 2 for the pattern. q3_2_a <- q3_2_b <- q3_2_c <- # Q3.3 ---- # 1 point # How many of the three race and ethnicity variables from Q3.2 had missing values? # Assign your answer to an object named `q3_3` as a numeric value. q3_3 <- # Q3.4 ---- # 1 point # Use `mutate()` to create a new variable `pct_bl_hisp_nat` in `df_race` that is the sum of # `pct_black`, `pct_hispanic`, and `pct_amerindian`. # Reminder: You are creating the new variable inside the `df_race` dataframe, not a new object, # so assign your result back to `df_race`. df_race <- # Q3.5 ---- # 1 point # Use `mutate()` to create a 0/1 indicator called `gt50pct_bl_hisp_nat` for whether more than 50% # of students identify as Black, Latinx, or Native American combined (i.e., `pct_bl_hisp_nat`) # Reminder: You are creating the new variable inside the `df_race` dataframe, not a new object. # HINT: "more than 50%" means use `>`, not `>=`. df_race <- # Q3.6 ---- # 1 point # Create frequency table using `count()` for the new analysis variable `gt50pct_bl_hisp_nat`. # Assign this table to a new object named `q3_6`. q3_6 <- # Q3.7 ---- # 1 point # Based on your analyses, what percent of schools in the `df_school_all` data # had a combined enrollment of Black, Latinx, and Native American students over 50%? # Assign your answer to an object named `q3_7` as a numeric value. # Please do not use any decimals or a percentage sign, just a rounded number from 0 to 100. # HINT: you can type the number directly, or calculate it with code using the counts in `q3_6` # and `round()`. q3_7 <- ## Question 4 ---- # Complete the following steps to create an analysis variable using `mutate()` and `case_when()`: # Q4.1 ---- # 1 point # First, use `select()` to select `name` and `state_code` from `df_school_all`. # Assign to a new dataframe named `df_schools`. df_schools <- # Q4.2 ---- # 2 points # Use `case_when()` to create a new variable in `df_schools` called `region`, with values: # 'Northeast' if state_code in c('CT','ME','MA','NH','RI','VT','NJ','NY','PA') # 'Midwest' if state_code in c('IN','IL','MI','OH','WI','IA','KS','MN','MO', # 'NE','ND','SD') # 'West' if state_code in c('AZ','CO','ID','NM','MT','UT','NV','WY','AK', # 'CA','HI','OR','WA') # 'South' if state_code is not any of the above states (use `.default`) # Assign the new variable creation back into the `df_schools` data frame # We did the above for you :) so just run these lines below. # NOTE: DC is not in any of the lists, so it is counted as South. That is expected. df_schools <- df_schools %>% mutate(region = case_when( state_code %in% c('CT','ME','MA','NH','RI','VT','NJ','NY','PA') ~ 'Northeast', state_code %in% c('IN','IL','MI','OH','WI','IA','KS','MN','MO','NE','ND','SD') ~ 'Midwest', state_code %in% c('AZ','CO','ID','NM','MT','UT','NV','WY','AK','CA','HI','OR','WA') ~ 'West', .default = 'South' )) # YOUR TURN: check the work above. Generate a frequency table of the new `region` variable # using the `count()` function. Assign this frequency table to an object named `q4_2`. q4_2 <- ## Complete the following steps to recode variables using `mutate()` and `recode_values()`: # Q4.3 ---- # 2 points # In the `df_schools` dataframe, replace the values of the `region` variable: # Northeast -> NE # Midwest -> MW # West -> W # South -> S # Remember to use assignment on `df_schools` to affect changes to the `region` variable. # Example code structure using `school_type` from `df_school_all`: # df_school_all %>% # mutate(school_type = recode_values(school_type, "public" ~ "PUB", "private" ~ "PRIV")) %>% # count(school_type) # Now do the same thing for `region` in `df_schools`, and assign the result back to `df_schools`! df_schools <- # Check your work by using the `count()` function on the updated `region` variable # Assign this frequency table to an object named `q4_3` q4_3 <- # Q4.4 ---- # 2 points # In the `df_schools` dataframe, create a new variable `state_name` whose value is: # 1. 'California' if state_code is 'CA' # 2. 'New York' if state_code is 'NY' # 3. One more state of your choice, e.g. 'Texas' if state_code is 'TX' (all 50 states plus DC are included) # 4. 'Other' for every school NOT in California, New York, or your chosen state (this is the "default" value -- see the hint below) # You may use either `case_when()` or `recode_values()`. # Do not change `state_code` itself. Create `state_name` as a new variable and assign the # result back to `df_schools`. # HINT: Both functions need a default for "any other state", but it is spelled differently: # case_when(..., .default = "Other") <- WITH a dot # recode_values(..., default = "Other") <- NO dot # The default goes after your last line and uses `=` (not `~`). df_schools <- ## Create a GitHub issue ---- # 2 points # Go to the class repository https://github.com/anyone-can-cook/rclass1_student_issues_f26/issues and create a new issue. # Refer to rclass1 student issues readme https://github.com/anyone-can-cook/rclass1_student_issues_f26/blob/main/README.md # for instructions on how to post questions or reflections. # You are also required to respond to at least one issue posted by another student. # Create an object named `issue` below and assign it the URL to your new issue, as a character object. # Make sure the URL is in quotes. issue <- # Create an object named `reply` below and assign it the URL to your response to another issue, as a character object. # Make sure the URL is in quotes. # HINT: To get a link to your specific comment, click the "..." menu at the top right of your # comment and choose "Copy link". reply <- ## Submit problem set on Canvas ---- # Use this naming convention "lastname_firstname_problemset#" for your R script (e.g. jaquette_ozan_problemset4.R).