exercise_solutions.md

0. Ensure tidyverse, dplyr is installed and loaded:

install.packages("tidyverse")
library("tidyverse")
install.packages("dplyr")
library("dplyr")

CCHS Data:

1. Read in the data cchs.csv

cchs <- read_csv("csv_file_location/cchs.csv")

2. Quickly view the first 10 rows of data

head(cchs, 10)

3. Display the type of variable for all variables in the dataset

sapply(cchs, class)

4. Create a scatterplot on height and weight

ggplot(cchs, aes(x = hwtghtm, y = hwtgwtk)) + geom_point()

5. Clean the height and weight data so missing data coded as numbers become NA in R

# Worse
cchs_clean <- cchs %>% mutate(hwtghtm = factor(hwtghtm))
cchs_clean <- cchs_clean %>% mutate(hwtgwtk = factor(hwtgwtk))
cchs_clean <- cchs_clean %>% mutate(hwtghtm = fct_recode(hwtghtm, NULL = "9.996", NULL = "9.999"))

# Better
cchs_clean <- cchs %>% mutate(hwtghtm = case_when(
	hwtghtm > 8.0 ~ NA_real_,
	TRUE ~ hwtghtm
))
cchs_clean <- cchs_clean %>% mutate(hwtgwtk = case_when(
	hwtgwtk > 900.0 ~ NA_real_,
	TRUE ~ hwtgwtk
))

6. Create a new scatterplot on height and weight with the clean data

ggplot(cchs_clean, aes(x = hwtghtm, y = hwtgwtk)) + geom_point()

7. Recode BMI to represent weight categories from underweight to obese

cchs_clean <- cchs_clean %>%
	mutate(bmi_category = case_when(
		hwtgbmi < 18.5 ~ "underweight",
		hwtgbmi >=30 & hwtgbmi <999 ~ "obese",
		hwtgbmi >=25 & hwtgbmi <30 ~ "overweight",
		hwtgbmi >=18.5 & hwtgbmi <25 ~ "normal weight",
		TRUE ~ "other"
	))

8. Recode Province to include the province names instead of numbers

cchs_clean <- cchs_clean %>%
	mutate(geogprv_name = case_when(
		geogprv == 10 ~ "NFLD & LAB",
		geogprv == 11 ~ "PEI",
		geogprv == 12 ~ "NOVA SCOTIA",
		geogprv == 13 ~ "NEW BRUNSWICK",
		geogprv == 24 ~ "QUEBEC",
		geogprv == 35 ~ "ONTARIO",
		geogprv == 46 ~ "MANITOBA",
		geogprv == 47 ~ "SASKATCHEWAN",
		geogprv == 48 ~ "ALBERTA",
		geogprv == 59 ~ "BRITISH COLUMBIA",
		geogprv == 60 ~ "YUKON/NWT/NUNA",
		geogprv == 96 ~ "NOT APPLICABLE",
		geogprv == 97 ~ "DON'T KNOW",
		geogprv == 98 ~ "REFUSAL",
		TRUE ~ "NOT STATED"
	))

9. Compute the mean and standard deviation of height and weight

summarize(
	cchs_clean, 
	avg_ht = mean(hwtghtm, na.rm = T), 
	sd_ht = sd(hwtghtm, na.rm = T),
	avg_wt = mean(hwtgwtk, na.rm = T), 
	sd_wt = sd(hwtgwtk, na.rm = T)
	)

10. Compute the mean and standard deviation of height and weight

summarize(
	group_by(cchs_clean, bmi), 
	avg_ht = mean(hwtghtm, na.rm = T), 
	sd_ht = sd(hwtghtm, na.rm = T),
	avg_wt = mean(hwtgwtk, na.rm = T), 
	sd_wt = sd(hwtgwtk, na.rm = T)
	)
summarize(
	group_by(cchs_clean, geogprv_name), 
	avg_ht = mean(hwtghtm, na.rm = T), 
	sd_ht = sd(hwtghtm, na.rm = T),
	avg_wt = mean(hwtgwtk, na.rm = T), 
	sd_wt = sd(hwtgwtk, na.rm = T)
	)

Accel Data

1. Read in the 2 data files accel.csv accel = read_csv("onboard/accel_data.csv")

2. Create a new variable that indicates participant 1 and participant 2

3. Append (stack) the 2 files together

4. Quickly view the first 10 rows of data

5. Display the type of variable for all variables in the dataset

6. Create a scatterplot on x_axis and y_axis

7. Compute the mean and standard deviation of height and weight

8. Convert the time data to time format

9. Compute the sum each of axis by second and by participant

10. Compute the gravity subtracted vector magnitude sqrt(x^2, y^2, z^2)-1 on the new data for each participant

Database

2. Open the connection to the interact_demo database, then use tbl() to get a reference to each of the tables in it.

drv <- dbDriver("PostgreSQL")
con <- dbConnect(drv, dbname = "interact_demo",
    host = "yakitori.usask.ca", port = 5432,
    user = rstudioapi::askForPassword("Database user"), 
    password = rstudioapi::askForPassword("Database password"))
demo_accel <- tbl(con, "accel_data")
demo_cchs <- tbl(con, "cchs")

3. View the first ten rows of each table.

demo_accel %>% head(10)
demo_cchs %>% head(10)

4. Display rows from the cchs table where the data originates in Saskatchewan.

demo_cchs %>% filter(geogprv == 47)

5. Display only the caseid column from the cchs table, where the data originates from Ontario

demo_cchs %>% filter(geogprv == 35) %>% select(caseid)

6. Display the average BMI by province from the cchs table

demo_cchs %>% filter(hwtgbmi < 50) %>%
	group_by(geogprv) %>%
	summarise(avg_bmi = mean(hwtgbmi))