Lab: Clean the air-quality file together
Reference solution for the Chapter 6 Practice Demo Lab (Tasks 1–6 of Lab: Clean the air-quality file together). Check your own attempt against it. The graded Canvas lab has its own private answer key.
air <- read_csv(here("data", "airdata3.csv"), show_col_types = FALSE)
dim(air) # rows x columns[1] 11904 27n_all_na <- sum(map_lgl(air, ~ all(is.na(.))))
n_with_data <- ncol(air) - n_all_na
c(columns = ncol(air), all_NA_columns = n_all_na, columns_with_data = n_with_data) columns all_NA_columns columns_with_data
27 14 13 The file has 11904 rows × 27 columns; 14 columns are 100 % missing and 13 carry data. glimpse(air) shows the key columns: location (the monitoring station), Date, Time, and pollutant columns including O3.
Drop the two spreadsheet index columns and every 100 %-missing column:
air2 <- air |>
select(-`...1`, -`Unnamed: 0`) |> # drop the spreadsheet index columns
select(where(~ !all(is.na(.)))) # drop columns that are entirely NA
ncol(air2) # columns remaining[1] 11names(air2) [1] "PM10" "SO2" "CO" "NO" "NO2" "PM25"
[7] "O3" "location" "Date" "Time" "TRS" That leaves the 11 columns that actually contain data — location, Date, Time, O3, and the other measured pollutants.
air3 <- air2 |> mutate(date = mdy(Date))
sum(is.na(air3$date)) # how many dates failed to parse?[1] 0mdy(Date) parses every row — the failure count is 0 — because the file stores dates as month/day/year (e.g., 1/1/2000). If any had failed, they would show up as NA in this count.
by_station <- air3 |>
filter(!is.na(O3)) |>
group_by(location) |>
summarise(mean_o3 = round(mean(O3), 1),
n_hours = n(),
.groups = "drop") |>
arrange(desc(mean_o3))
kable(by_station, caption = "Mean hourly ozone (O3, ppb) by station, January 2000, highest first.")| location | mean_o3 | n_hours |
|---|---|---|
| Langley Central | 17.7 | 732 |
| Maple Ridge Golden Ears School | 11.3 | 564 |
| Richmond South | 10.5 | 732 |
| North Delta | 10.3 | 731 |
| Vancouver International Airport #2 | 10.2 | 727 |
| Burnaby South | 9.1 | 732 |
| Vancouver Kitsilano | 7.5 | 731 |
| Port Moody Rocky Point Park | 6.1 | 728 |
Langley Central has the highest mean O₃ (17.7 ppb) for the month. That is consistent with Metro Vancouver geography: ozone is a secondary pollutant that forms downwind of traffic emissions, so cleaner-air / more suburban or elevated sites often record higher ozone than the busiest downtown stations — the opposite of what many students expect.
Each group submits: the final pipeline (Tasks 2–4 as one Quarto chunk), the station-ranked table from Task 4, and a one-sentence interpretation (Task 5), with all members’ names. (For the graded Canvas lab, follow the Canvas instructions.)