Parse Output from 'BabyTime' Application

'BabyTime' is an application for tracking infant and toddler care activities like sleeping, eating, etc. This package will take the outputted .zip files and parse it into a usable list object with cleaned data. It handles malformed and incomplete data gracefully and is designed to parse one directory at a time.

babyTimeR

The goal of babyTimeR is to parse the output from the BabyTime App into usable tibbles for analyzing sleeping and eating habits.

Installation

You can install the development version of babyTimeR from GitHub with:

# install.packages("pak")
pak::pak("dannhek/babyTimeR")

Example Sleep Graph

Example of tracking sleep over time

library(babyTimeR)
if (file.exists('baby_db.RDS')) {
	baby_db <- readRDS('baby_db.RDS')
} else {
	baby_db <- process_one_directory(directory = 'Raw Data')
	saveRDS(baby_db, 'baby_db.RDS')
}

library(tidyverse)
sleep_df <- baby_db$sleep |>
	mutate(
		sleep_night = case_when(
			hour(start_dttm) <= 6 ~ glue("{format(start_dttm-days(1),'%Y%m%d')}-{format(start_dttm,'%Y%m%d')}"),
			hour(start_dttm) >= 19 ~ glue("{format(start_dttm,'%Y%m%d')}-{format(start_dttm+days(1),'%Y%m%d')}"),
			.default = 'Daytime Nap'
		) 
	) |>
	select(-baby_name,-memo,-type) |>
	distinct()

gg_sleep_df <- sleep_df |>
	filter(sleep_night != 'Daytime Nap') |>
	arrange(start_dttm) |> 
	group_by(sleep_night) |> 
	mutate(
		time_awake = difftime(start_dttm, lag(end_dttm), units = 'mins')
	) |> 
	summarise(
		n_sleep_entries = n(), 
		longest_sleep = max_na(duration_min), 
		wake_ups = sum(time_awake > minutes(15), na.rm = T),
		longest_wake_up = max_na(time_awake)
	) 
	
glimpse(gg_sleep_df)

colors <- RColorBrewer::brewer.pal(name = 'Set1', n = 3)
ratio <- 60

ggplot(
	gg_sleep_df |> drop_na(), 
	aes(x = ymd(substr(sleep_night,1,8)))
) +
	# Sleep Length
	geom_bar(
		aes(
			fill = 'Longest Sleep of Night', 
			y = longest_sleep
		), 
		stat = 'identity', 
		alpha = 0.4
	) +
	geom_line(aes(
		color = 'Moving Average (Longest Sleep)', 
		y = zoo::rollmean(longest_sleep, 7, fill = NA, na.rm = T)
	)) +
	# Sleep Count
	geom_line(aes(
		color = 'Number of Wake-ups', 
		y = ratio*n_sleep_entries
	)) +
	# Wake Length
	geom_point(aes(
		shape = 'Longest Wake Period', 
		y = as.numeric(longest_wake_up)
	)) +
	# Scales 
	scale_x_date(
		name = 'Date', 
		date_breaks = '3 weeks', 
		date_labels = '%d%b%y'
	) +
	scale_y_continuous(
		name = 'Longest Sleep', 
		labels = display_time, 
		limits = c(0,10*ratio), 
		breaks = ratio*seq(0.5,9.5,1),
		sec.axis = sec_axis(
			~./ratio,name = 'Number of Wake-ups', 
			breaks = c(0:10)
		)
	) +
	scale_color_manual(
		name = '', 
		values = colors[2:3]
	) +
	scale_fill_manual(
		name = '', 
		values = colors[1]
	) +
	scale_shape(name = '') +
	# Theme/Look'n'Feel
	theme_minimal() +
	theme(
		legend.position = 'bottom',
		axis.text.x = element_text(angle = 60, vjust = 1, hjust = 1),
		plot.title.position = 'plot',
		plot.caption.position = 'plot'
	) +
	labs(
		title = 'Overnight Sleep Trends for My Baby',
		subtitle = 'Includes sleeps recorded as beginning between 7p and 7a',
		caption = 'Data collected via BabyTime app'
	)