## Problemset Title ---- ## [ PROJ ] < Problem set 2 SOLUTIONS > ## [ FILE ] < problemset2_solutions.R > ## [ AUTH ] < Teaching team > ## Notes ---- # - Where there is more than one correct way to answer, the options are separated by `# or`. # - Expected values are given in comments rm(list = ls()) # remove all objects from the Global Environment ## Question 1: Investigating objects, Base R ---- # Q1.1 ---- load(url("https://github.com/ozanj/rclass/raw/master/data/recruiting/recruit_school_allvars.RData")) # `url()` creates a connection to the web address, and `load()` reads the .RData file through # it. `load()` does not need to be assigned to anything: it puts `df_school_all` straight into # the Global Environment. Check: dim(df_school_all) is 21301 rows x 55 columns typeof(df_school_all) q1_x <- typeof(df_school_all) # option 1 is to assign your code to the object q1_x <- "list" # option 2 is to assign the literal output from typeof() to the object # A data frame is stored as a list, so typeof() returns "list" (class() would return # "data.frame"; for a tibble, "tbl_df" "tbl" "data.frame"). # Q1.2 ---- q1_2a <- length(df_school_all) # or q1_2a <- 55 q1_2b <- "variables" # or q1_2b <- "columns" # 55. A data frame is a list where each element is one column, so length() counts columns. # The test also accepts "elements". # Q1.3 ---- q1_3a <- nrow(df_school_all) # or q1_3a <- 21301 q1_3b <- "observation" # 21301. Each row is one observation, here one high school. # Q1.4 ---- q1_4 <- "variable" # Each element of a data frame is one variable (column). Note: the test only accepts # "variable"; a student answering "column" is also conceptually right. # Q1.5 ---- q1_5 <- "vectors" # Each column is an atomic vector. str() shows this with labels like `chr` and `num` # for each column. # Q1.6 ---- q1_6 <- "named" # Every column has a name (state_code, school_type, ...), which is what str() lists # after each `$`. # Q1.7 ---- q1_7 <- typeof(df_school_all$school_type) # or q1_7 <- typeof(df_school_all[["school_type"]]) # "character" # Q1.8 ---- q1_8 <- length(df_school_all$school_type) # or q1_8 <- length(df_school_all[["school_type"]]) # 21301, the same as the number of rows, because each column has one value per row. # Q1.9 ---- q1_9 <- "observations" # or q1_9 <- "rows" # For a single vector, length() counts its elements; for a column of a data frame, each # element is one row/observation. Note: the test accepts "row" or "observation" but not # "elements", which is also technically correct. # Q1.10 ---- q1_10_1 <- table(df_school_all$school_type) q1_10_2 <- table(df_school_all$school_type, useNA = "ifany") q1_10_3 <- table(df_school_all$school_type, useNA = "always") # private 3822, public 17479 in all three. # "ifany" adds an NA column only if there are missing values. `school_type` has none, # so the table looks the same as the default. # "always" adds an NA column no matter what, so the third table shows NA = 0. # Q1.11 ---- q1_11 <- "no" # From ?table: useNA = c("no", "ifany", "always"). The first option listed is the default. # Q1.12 ---- q1_12 <- 2 # private and public only, because there are no missing values. # Q1.13 ---- q1_13 <- 3 # private, public, and NA (with a count of 0). ## Question 2: Subsetting, Base R ---- # Create a named numeric atomic vector vec <- c(a = 2.4, b = 1.1, c = 3.4, d = 4, e = 6, f = 32, g = 21, h = 17, i = 10) str(vec) # Create a list list <- list(c(1:3), list("red", "orange"), list("LA", "NY", "DC")) str(list) # View the `df_school_all` data frame you loaded earlier head(df_school_all, n = 5) # Q2.1 ---- q2_1 <- vec[c(4, 7)] # or q2_1 <- vec[c("d", "g")] # d = 4, g = 21. `c()` is needed to give `[]` more than one position at a time. # Q2.2 ---- q2_2 <- vec[-9] # or q2_2 <- vec[-length(vec)] # or q2_2 <- vec[1:8] # a through h. A negative index removes that position. `-length(vec)` is the most general # because it works no matter how long the vector is. # Q2.3 ---- q2_3 <- vec[c("a", "d", "g")] # a = 2.4, d = 4, g = 21. Named vectors can be subset with a character vector of names. # Q2.4 ---- q2_4 <- vec[vec < 12] # a, b, c, d, e, i (2.4, 1.1, 3.4, 4, 6, 10). # `vec < 12` creates a logical vector (TRUE/FALSE for each element), and `[]` keeps the # elements that are TRUE. # Q2.5 ---- q2_5 <- list[1] # A list of length 1 containing the vector 1 2 3. # Q2.6 ---- q2_6 <- "list" # or q2_6 <- typeof(list[1]) # Single brackets `[]` on a list always return a list (a smaller piece of the same container). # Q2.7 ---- q2_7 <- df_school_all[c(2, 4, 6)] # or q2_7 <- df_school_all[, c(2, 4, 6)] # The columns school_type, name, and city. With one index and no comma, `[]` treats the # data frame as a list and returns columns. With a comma, it is [rows, columns], and leaving # the rows blank means "all rows." Both return a data frame with 3 columns. # Q2.8 ---- q2_8 <- df_school_all[1:100, "state_code"] # or q2_8 <- df_school_all$state_code[1:100] # or q2_8 <- df_school_all[["state_code"]][1:100] # Rows go before the comma, columns after: df[rows, columns]. # The `$` and `[[]]` options return a character vector of length 100. The first option returns # a 100 x 1 tibble if df_school_all is a tibble (a plain data.frame gives a vector instead). # The test accepts either. # Q2.9 ---- q2_9 <- df_school_all[1:3, c("state_code", "name")] # 3 x 2: AK "Bethel Regional High School", AK "Ayagina'ar Elitnaurvik", # AK "Kwigillingok School". Column order must be state_code then name to match the test. # Q2.10 ---- q2_10 <- list[[1]] # The vector 1 2 3 itself, not wrapped in a list. # The usual analogy: if `list` is a train, `list[1]` is the first train car, and # `list[[1]]` is what's inside the first car. # Q2.11 ---- q2_11 <- "integer" # or q2_11 <- typeof(list[[1]]) # `1:3` creates integers, so the contents of the first element are an integer vector. # Q2.12 ---- q2_12 <- "list" # or q2_12 <- typeof(df_school_all["total_students"]) # `[]` returns a data frame with one column, which is stored as a list. # Q2.13 ---- q2_13 <- "double" # or q2_13 <- typeof(df_school_all[["total_students"]]) # `[[]]` pulls the column out as a vector, which is numeric (double). # Q2.14 ---- q2_14 <- "double" # or q2_14 <- typeof(df_school_all$total_students) # `$` works the same as `[[]]` with a name, so it also returns the double vector. # Q2.15 ---- q2_15 <- df_school_all$total_students[10:15] # 186 213 211 324 421 181. `$` extracts the vector first, then `[]` picks positions 10-15. ## Question 3: Calculate the total average number of students receiving free lunch from `num_fr_lunch`. # Q3.1 - q3.3 ---- q3_1 <- mean(df_school_all["num_fr_lunch"], na.rm = TRUE) q3_2 <- mean(df_school_all[["num_fr_lunch"]], na.rm = TRUE) q3_3 <- mean(df_school_all$num_fr_lunch, na.rm = TRUE) # q3_1 returns NA with a warning ("argument is not numeric or logical: returning NA"), # because `[]` returns a one-column data frame (a list), and mean() can't average a list. # q3_2 and q3_3 both return about 322.79, because `[[]]` and `$` return the numeric vector. # `na.rm = TRUE` drops missing values before averaging; without it, the result would be NA. # Q3_4 ---- q3_4 <- 1 # Q3_5 ---- q3_5 <- "list" # Q3_6 ---- q3_6 <- c("logical", "numeric") # From ?mean, x must be a numeric or logical vector. "numeric" includes both integer and # double. Also accepts c("double", "integer", "logical"). # Logical works because TRUE counts as 1 and FALSE as 0, so the mean of a logical vector # is the proportion of TRUEs. ## Create a GitHub issue ---- # Example format only. Each student's URLs will be different. # The test requires the issue URL to end in /issues/ and the reply URL to end in # /issues/#issuecomment- (get this from "..." > "Copy link" on the comment). issue <- "https://github.com/anyone-can-cook/rclass1_student_issues_f26/issues/1" reply <- "https://github.com/anyone-can-cook/rclass1_student_issues_f26/issues/2#issuecomment-1234567890"