The overall goal of this project is to use personal activity monitoring data of a step counter throughout the day in 5-minute intervals between October and November - 2012, and build a literate exploratory data analysis aimed at reproducibility.
- R language compiler
- R base graphic devices
- Tidyverse library packages
- RMarkdown library package
- Knitr library package
- Raw data set: Activity Monitoring Data
- CODEBOOK: step-by-step book explaining the code processing.
- Figures: the plotted images
- PA1_template.Rmd: the script to compile the project from source
Inside the project root folder, execute the following
R
library(rmarkdown)
rmarkdown::render("PA1_template.Rmd", "html_document")
library(tidyverse)## ── Attaching packages ───────────────────────────────────────────────────────────────────────────────── tidyverse 1.3.0 ──
## ✓ ggplot2 3.3.0 ✓ purrr 0.3.4
## ✓ tibble 3.0.1 ✓ dplyr 0.8.5
## ✓ tidyr 1.0.2 ✓ stringr 1.4.0
## ✓ readr 1.3.1 ✓ forcats 0.5.0
## ── Conflicts ──────────────────────────────────────────────────────────────────────────────────── tidyverse_conflicts() ──
## x dplyr::filter() masks stats::filter()
## x dplyr::lag() masks stats::lag()
library(RColorBrewer)
set.seed(144)# unzip("activity.zip")
# file.remove("activity.zip")activity <- read_csv("activity.csv")## Parsed with column specification:
## cols(
## steps = col_double(),
## date = col_date(format = ""),
## interval = col_double()
## )
glimpse(activity)## Rows: 17,568
## Columns: 3
## $ steps <dbl> NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, …
## $ date <date> 2012-10-01, 2012-10-01, 2012-10-01, 2012-10-01, 2012-10-01,…
## $ interval <dbl> 0, 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 55, 100, 105, 110,…
Count number of NAs
activity %>%
summarize_all(funs(sum(is.na(.))))## Warning: funs() is soft deprecated as of dplyr 0.8.0
## Please use a list of either functions or lambdas:
##
## # Simple named list:
## list(mean = mean, median = median)
##
## # Auto named with `tibble::lst()`:
## tibble::lst(mean, median)
##
## # Using lambdas
## list(~ mean(., trim = .2), ~ median(., na.rm = TRUE))
## This warning is displayed once per session.
## # A tibble: 1 x 3
## steps date interval
## <int> <int> <int>
## 1 2304 0 0
mean_steps_day <- activity %>%
group_by(date) %>%
drop_na() %>%
summarize(
mean.steps = mean(steps),
interval = sum(interval)
)
mean_steps_day## # A tibble: 53 x 3
## date mean.steps interval
## <date> <dbl> <dbl>
## 1 2012-10-02 0.438 339120
## 2 2012-10-03 39.4 339120
## 3 2012-10-04 42.1 339120
## 4 2012-10-05 46.2 339120
## 5 2012-10-06 53.5 339120
## 6 2012-10-07 38.2 339120
## 7 2012-10-09 44.5 339120
## 8 2012-10-10 34.4 339120
## 9 2012-10-11 35.8 339120
## 10 2012-10-12 60.4 339120
## # … with 43 more rows
ggplot(mean_steps_day, aes(mean.steps)) +
geom_histogram(bins = 30) +
labs(
title = "Sum of mean steps",
x = "Steps taken",
y = "Count"
) +
theme_minimal()
ggplot(mean_steps_day, aes(date, mean.steps)) +
geom_bar(stat = "identity") +
labs(
title = "Mean of steps",
x = "Date",
y = "Mean steps"
) +
theme_minimal()
median_steps_day <- activity %>%
group_by(date) %>%
drop_na() %>%
filter(steps != 0) %>%
summarize(
median.steps = median(steps),
interval = sum(interval)
)
median_steps_day## # A tibble: 53 x 3
## date median.steps interval
## <date> <dbl> <dbl>
## 1 2012-10-02 63 4425
## 2 2012-10-03 61 97695
## 3 2012-10-04 56.5 125300
## 4 2012-10-05 66 95090
## 5 2012-10-06 67 132435
## 6 2012-10-07 52.5 138005
## 7 2012-10-09 48 118665
## 8 2012-10-10 56.5 132915
## 9 2012-10-11 35 92745
## 10 2012-10-12 46 138400
## # … with 43 more rows
ggplot(median_steps_day, aes(median.steps)) +
geom_histogram(bins = 30) +
labs(
title = "Sum of median steps",
x = "Steps taken",
y = "Count"
) +
theme_minimal()
ggplot(median_steps_day, aes(date, median.steps)) +
geom_bar(stat = "identity") +
labs(
title = "Median of steps",
x = "Date",
y = "Median steps"
) +
theme_minimal()
mean_interval <- activity %>%
drop_na() %>%
group_by(interval) %>%
summarize(
mean.steps = mean(steps)
)
mean_interval## # A tibble: 288 x 2
## interval mean.steps
## <dbl> <dbl>
## 1 0 1.72
## 2 5 0.340
## 3 10 0.132
## 4 15 0.151
## 5 20 0.0755
## 6 25 2.09
## 7 30 0.528
## 8 35 0.868
## 9 40 0
## 10 45 1.47
## # … with 278 more rows
ggplot(mean_interval, aes(interval, mean.steps)) +
geom_line() +
labs(
title = "Average daily activity",
x = "Time (minutes)",
y = "Global mean steps",
color = ""
) +
theme_minimal()
Which 5-minute interval, on average across all the days in the dataset, contains the maximum number of steps?
mean_interval %>%
group_by(interval) %>%
arrange(desc(mean.steps))## # A tibble: 288 x 2
## # Groups: interval [288]
## interval mean.steps
## <dbl> <dbl>
## 1 835 206.
## 2 840 196.
## 3 850 183.
## 4 845 180.
## 5 830 177.
## 6 820 171.
## 7 855 167.
## 8 815 158.
## 9 825 155.
## 10 900 143.
## # … with 278 more rows
Compare means
mean_act <- mean(activity$steps, na.rm = TRUE)
mean_act## [1] 37.3826
Imputing mean of that day on missing values
iactivity <- activity %>%
mutate(steps = ifelse(is.na(steps), 0, steps)) %>%
group_by(date) %>%
mutate(
steps = ifelse(steps==0, mean(steps), steps)
) %>%
ungroup()
mean_iact <- mean(iactivity$steps)
mean_iact## [1] 55.17414
Mean of 37.3825996 steps for NAs removed.
Mean of 55.1741376 steps for imputed mean.
Make a histogram of the total number of steps taken each day and Calculate and report the mean and median total number of steps taken per day. Do these values differ from the estimates from the first part of the assignment?
imean_steps_day <- iactivity %>%
group_by(date) %>%
summarize(
mean.steps = mean(steps),
interval = sum(interval)
)
imean_steps_day## # A tibble: 61 x 3
## date mean.steps interval
## <date> <dbl> <dbl>
## 1 2012-10-01 0 339120
## 2 2012-10-02 0.872 339120
## 3 2012-10-03 67.7 339120
## 4 2012-10-04 69.5 339120
## 5 2012-10-05 78.5 339120
## 6 2012-10-06 87.4 339120
## 7 2012-10-07 62.2 339120
## 8 2012-10-08 0 339120
## 9 2012-10-09 74.3 339120
## 10 2012-10-10 56.3 339120
## # … with 51 more rows
ggplot(imean_steps_day, aes(mean.steps)) +
geom_histogram(bins = 30) +
labs(
title = "Sum of mean steps",
x = "Steps taken",
y = "Count"
) +
theme_minimal()
ggplot(imean_steps_day, aes(date, mean.steps)) +
geom_bar(stat = "identity") +
labs(
title = "Mean of steps",
x = "Date",
y = "Mean steps"
) +
theme_minimal()
imedian_steps_day <- iactivity %>%
group_by(date) %>%
summarize(
median.steps = median(steps),
interval = sum(interval)
)
imedian_steps_day## # A tibble: 61 x 3
## date median.steps interval
## <date> <dbl> <dbl>
## 1 2012-10-01 0 339120
## 2 2012-10-02 0.438 339120
## 3 2012-10-03 39.4 339120
## 4 2012-10-04 42.1 339120
## 5 2012-10-05 46.2 339120
## 6 2012-10-06 53.5 339120
## 7 2012-10-07 38.2 339120
## 8 2012-10-08 0 339120
## 9 2012-10-09 44.5 339120
## 10 2012-10-10 34.4 339120
## # … with 51 more rows
ggplot(imedian_steps_day, aes(median.steps)) +
geom_histogram(bins = 30) +
labs(
title = "Sum of median steps",
x = "Steps taken",
y = "Count"
) +
theme_minimal()
ggplot(imedian_steps_day, aes(date, median.steps)) +
geom_bar(stat = "identity") +
labs(
title = "Median of steps",
x = "Date",
y = "Median steps"
) +
theme_minimal()
Inserting imputed values allowed to remove drop_na and to remove filter !=0 and still get significant median values.
Joining datasets and plotting mean differences
# Checking dimensions
dim(mean_steps_day)## [1] 53 3
dim(imean_steps_day)## [1] 61 3
# Joining datasets and pivoting into one mean column
join_mean <- left_join(imean_steps_day, mean_steps_day, by = c("date", "interval"))
meta <- pivot_longer(data = join_mean, c(`mean.steps.x`, `mean.steps.y`), names_to = "group", values_to="mean.steps")
group_mean <- meta %>%
mutate(
group = ifelse(group == "mean.steps.x", "imean", "mean")
) %>%
drop_na()
# Plotting results
ggplot(group_mean, aes(date, mean.steps, group = group, fill = group)) +
geom_bar(stat = "identity") +
labs(
title = "Mean difference between imputed and non-imputed datasets",
x = "Date",
y = "Mean of steps"
) +
theme_minimal()
The imputed data received considerable mean increase.
day <- weekdays(activity$date)
daylevel <- vector()
for (i in 1:nrow(activity)) {
if (day[i] == "Saturday") {
daylevel[i] <- "Weekend"
} else if (day[i] == "Sunday") {
daylevel[i] <- "Weekend"
} else {
daylevel[i] <- "Weekday"
}
}
activity$daylevel <- daylevel
activity$daylevel <- factor(activity$daylevel)
dailysteps <- aggregate(steps ~ interval + daylevel, data = activity, mean)
names(dailysteps) <- c("interval", "daylevel", "steps")
glimpse(dailysteps)## Rows: 576
## Columns: 3
## $ interval <dbl> 0, 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 55, 100, 105, 110,…
## $ daylevel <fct> Weekday, Weekday, Weekday, Weekday, Weekday, Weekday, Weekda…
## $ steps <dbl> 2.3333333, 0.4615385, 0.1794872, 0.2051282, 0.1025641, 1.512…
ggplot(dailysteps, aes(interval, steps, group = daylevel, color = daylevel)) +
facet_grid(daylevel ~ .) +
geom_line() +
labs(
x = "Time (mins)",
y = "Steps",
color = "Type"
) +
theme_bw()