Nothing
R/Person.R
In lifelogr: Life Logging
#' @include global_var.R
NULL
#' A \code{Person} object is a complete view of an individual over a certain
#' time period, as seen through data from multiple sources
#'
#' @description \code{Person} is an object that encapsulates an individual's
#' data over a specified date range (start and end date stored as \code{Date}
#' objects.
#' An individual consists of basic information, such as name, age,
#' and gender (a \code{list} with named elements), data from their self-tracking
#' devices such as Fitbit, Apple health, etc. (data from each source is a tibble
#' dataframe), individual goals such as target steps (\code{numeric}),
#' additional data from self-tracking apps or one's own collection system
#' (stored as a tibble dataframe), and ways of grouping the data a user may be
#' interested in, such as grouping by seasons, or comparing weekend to weekday
#' behavior and health (stored as a list of named dataframes, which each contain
#' group assignments).
#'
#' @docType class
#' @importFrom R6 R6Class
#' @importFrom tibble data_frame as_data_frame
#' @importFrom lubridate hour minute second make_datetime ymd wday month
#' @importFrom fitbitScraper login get_intraday_data get_weight_data
#' get_daily_data get_sleep_data
#' @importFrom dplyr select mutate
#' @importFrom plyr rename
#' @field fitbit_daily tibble dataframe of fitbit variables (for user account
#' info provided) observed daily.
#' Columns include:
#' \itemize{
#' \item{date: unique for each row (date)}
#' \item{datetime: includes date and time, time is an arbitrary
#' time, which is consistent for each day (date)}
#' \item{dateInForJavascriptLocalFormatting: chr}
#' \item{steps: total number of steps for that day (dbl)}
#' \item{distance: total distance for that day, in miles (dbl)}
#' \item{distanceKm: total distance for that day, in kilometers
#' (dbl)}
#' \item{floors: total number of floors for that day (dbl)}
#' \item{minutesVery: minutes 'very active' that day (dbl)}
#' \item{caloriesBurned: calories (kcal) burned that day (dbl)}
#' \item{caloriesIntake: calories (kcal) consumed that day, user
#' must input this, either into this data frame or into the fitbit
#' (dbl)}
#' \item{restingHeartRate: resting heart rate in beats per minute
#' (bpm) (dbl)}
#' \item{startTime: sleeping start time for that day (chr)}
#' \item{endTime: sleeping end time for that day (chr)}
#' \item{startDateTime: sleeping start date and time for that day
#' (chr)}
#' \item{endDateTime: sleeping end date and time for that day (chr)}
#' \item{sleepDuration: sleep duration for that day, in minutes
#' (int)}
#' \item{sleepDurationHrs: sleep duration for that day, in hours
#' (dbl)}
#' \item{minAsleep: time asleep that day, in minutes (int)}
#' \item{minAsleepHrs: time asleep that day, in hours, derived from
#' minAsleep (dbl)}
#' \item{minRestlessAwake: \eqn{sleepDuration - awakeCount} (int)}
#' \item{awakeCount: int}
#' \item{restlessCount: int}
#' \item{awakeDuration: int}
#' \item{restlessDuraton: int}
#' \item{restlessProp: proportion of sleep spent restless,
#' calculated as \eqn{restlessProp = (sleepDurationHrs -
#' minAsleepHrs) / sleepDurationHrs * 100} (dbl)}
#' \item{sleepQualityScoreB: dbl}
#' \item{sleepQualityScoreA: int}
#' \item{sleepQualityGraphicPercentA: dbl}
#' \item{sleepQualityGraphicPercentB: dbl}
#' \item{sleepBucketTextB: one of "ok", "good", "great" (chr)}
#' \item{sleepBucketTextA: one of "ok", "good", "great" (chr)}
#' \item{clusters: list of chr}
#' \item{breaks: list of chr}
#' }
#' @field fitbit_intraday tibble dataframe of fitbit variables (for user account
#' info provided) observed multiple times a day.
#' Columns include:
#' \itemize{
#' \item{date: unique for each row (date)}
#' \item{time: a combination of an arbitrary date ("1970-01-01")
#' and the time of the observation, generally in 5 minute intervals
#' (dttm)}
#' \item{datetime: includes date from `date` and time from `time`
#' (dttm)}
#' \item{steps: number of steps in 15 minute interval (dbl)}
#' \item{distance: distance traveled in 15 minute interval, in
#' miles (dbl)}
#' \item{distanceKm: distance traveled in 15 minute interval, in
#' kilometers (dbl)}
#' \item{floors: number of floors went up and down in 15 minute
#' interval (dbl)}
#' \item{activeMin: number of active minutes in 15 minute
#' interval (dbl)}
#' \item{activityLevel: hypothesized activity level, one of:
#' "SEDENTARY", "LIGHTLY_ACTIVE", "MODERATELY_ACTIVE", or
#' "VERY_ACTIVE" (chr)}
#' \item{bpm: average heart rate in 5 minute interval (int)}
#' \item{confidence: one of -1, 1, 2, or 3 (int)}
#' \item{caloriesBurned: calories (kcal) burned in 5 minute
#' interval (dbl)}
#' \item{defaultZone: chr}
#' \item{customZone: lgl}
#' \item{weight: weight, in lbs (dbl)}
#' \item{weightKg: weight, in kg (dbl)}
#' }
#' @field util tibble dataframe that maps each date in the date range to utility
#' information about that date
#' Columns include:
#' \itemize{
#' \item{date: unique for each row (date)}
#' \item{datetime: date from `date` and an arbitrary time (16:00:00)
#' (dttm)}
#' \item{day_of_week: day of the week, with Sun as first (ord)}
#' \item{day_type: weekend or weekday (fctr)}
#' \item{month: month, with Jan as first (ord)}
#' }
#' @field target_steps the person's target number of steps (numeric) for each
#' day (default 10,000)
#' @field start_date start of user's date range of interest (Date object)
#' @field end_date end of user's date range of interest (Date object)
#' @field user_info provided user info, such as "age", "gender", "name" (list)
#' @field groupings named list of dataframes, each with two columns - a known
#' variable, and group, with the group assignment for observations where
#' that variable has appropriate value
#' @field apple tibble dataframe of user's provided Apple Health data. These
#' columns depend on which columns are passed in by the user. However,
#' these columns match fitbit columns:
#' \itemize{
#' \item{datetime: dttm}
#' \item{steps: Original steps data for total number of steps in 60
#' minutes, but divided by 4 to match fitbit steps data, which is
#' the total number of steps in 15 minutes (dbl)}
#' \item{distance: Average distance in 15 minutes in miles.
#' Original distance data for total distance in 60 minutes,
#' but divided by 4 to match fitbit distance data, which is the total
#' distance in 15 minutes (dbl)}
#' \item{distanceKm: Average distance in 15 minutes in km.
#' Original distance data for total distance in 60 minutes,
#' but divided by 4 to match fitbit distance data, which is the total
#' distance in 15 minutes (dbl)}
#' \item{floors: Average number of floors in 15 minutes.
#' Original floors data for total number
#' of floors in 60 minutes, but divided by 4 to match fitbit floors
#' data, which is the total number of floors in 15 minutes(dbl)}
#' \item{bpm: average heart rate for the given hour, most users will
#' not have this data (dbl)}
#' }
#' @field addl_data dataframe of data from another source provided by user
#' @field addl_data2 dataframe of data from another source provided by user
#'
#' @section Methods:
#'
#' \describe{
#' \item{\code{Person$new(fitbit_user_email, fitbit_user_pw, user_info = NA,
#' apple_data_file, target_steps, addl_data,
#' addl_data2, group_assignments, start_date,
#' end_date)}}{Creates a new \code{Person}
#' with specified data, and data from provided Fitbit account. If
#' provided, start_date and end_date must be characters with "%Y-%m-%d".
#' All defaults are \code{NA} - user can provide sources of data
#' of interest.}}
#'
#' @examples
#' library("lifelogr")
#'
#' group_months <- data.frame("month" = c("Jan", "Feb", "Mar", "Apr", "May",
#' "Jun", "Jul", "Aug",
#' "Sep", "Oct", "Nov", "Dec"),
#' "group" = c(0, 0, 0, 1, 1, 1, 1, 1,
#' 1, 0, 0, 0))
#' ash <- Person$new(user_info = list("name" = "Ash", "age" = 26,
#' "gender" = "female"),
#' target_steps = 20000,
#' group_assignments = list("group_months" = group_months),
#' start_date = "2017-03-11",
#' end_date = "2017-03-12")
#'
#' \dontrun{
#' bailey <- Person$new(fitbit_user_email = "bailey@gmail.com",
#' fitbit_user_pw = "baileypw",
#' #apple_data_file = "apple.csv",
#' start_date = "2017-03-11",
#' end_date = "2017-03-12")}
#'
#' @export
#' @format An \code{\link{R6Class}} generator object
#'
Person <- R6::R6Class("Person",
public = list(
fitbit_daily = NULL, # dataframe of daily fitbit data
fitbit_intraday = NULL, # dataframe of intraday fitbit data
util = NULL, # util/admin data
target_steps = NULL,
start_date = NA, # optional start date of interest
end_date = NA, # optional end date of interest
user_info = NULL, # optional list of user info, such as "age", "gender", "name", etc.
groupings = NULL, # named list of dataframes each with 2 columns: 1. a known
# variable (date, weekend, etc.) and 2. "group" with the group assignment for
# observations with that variable of that value
apple = NULL, # apple data from csv
addl_data = NULL, # user's own df of other data
addl_data2 = NULL, # another df
initialize = function(fitbit_user_email = NA, fitbit_user_pw = NA,
user_info = NA,
apple_data_file = NA,
target_steps = 10000,
addl_data = NA, addl_data2 = NA,
group_assignments = NA,
start_date = NA, end_date = NA) {
self$addl_data <- addl_data
self$addl_data2 <- addl_data2
if (is.na(start_date)) {
start_date = "1990-01-01"
}
if (is.na(end_date)) {
end_date = Sys.Date()
}
self$start_date <- as.Date(strptime(start_date, format="%Y-%m-%d"))
self$end_date <- as.Date(strptime(end_date, format="%Y-%m-%d"))
self$target_steps <- target_steps
self$user_info <- user_info
self$util <- private$create_util_data(self$start_date, self$end_date)
if (!is.na(apple_data_file)){
self$apple <- private$load_apple_data(apple_data_file)
# user needs to pass in path from this directory to file
}
if (!is.na(fitbit_user_email) && !is.na(fitbit_user_pw)){
self$fitbit_intraday <- private$get_fitbit_intraday(fitbit_user_email,
fitbit_user_pw)
self$fitbit_daily <- private$get_fitbit_daily(fitbit_user_email,
fitbit_user_pw)
}
else {
self$fitbit_intraday <- NA
self$fitbit_daily <- NA
}
self$groupings <- group_assignments
}),
private = list(
load_apple_data = function(apple_data_file) {
raw_df <- readr::read_csv(apple_data_file)
cleaned <- tibble::as_tibble(raw_df)
# Remove Finish column because it's confusing
if(!is.null(cleaned$Finish)) {
cleaned$Finish <- NULL
}
# Rename columns to match Person$data_intraday and add additional columns
if("Start" %in% names(cleaned)) {
cleaned <- dplyr::rename(cleaned, datetime = Start)
}
# fitbit steps data is in 15 min intervals, while Apple steps data is in
# hour intervals
if("Steps (count)" %in% names(cleaned)) {
cleaned <- dplyr::rename(cleaned, steps = `Steps (count)`)
cleaned <- dplyr::mutate(cleaned,
steps = steps / 4)
}
# fitbit floors data is in 15 min intervals, while Apple floors data
# is in hour intervals
if("Flights Climbed (count)" %in% names(cleaned)) {
cleaned <- dplyr::rename(cleaned, floors = `Flights Climbed (count)`)
cleaned <- dplyr::mutate(cleaned, floors = floors / 4)
}
if("Heart Rate (count/min)" %in% names(cleaned)) {
cleaned <- dplyr::rename(cleaned, bpm = `Heart Rate (count/min)`)
}
# fitbit distance data is in 15 min intervals, while Apple distance data
# is in hour intervals
if("Distance (mi)" %in% names(cleaned)) {
cleaned <- dplyr::rename(cleaned, distance = `Distance (mi)`)
cleaned <- dplyr::mutate(cleaned,
distance = distance / 4,
distanceKm = distance * MI_TO_KM)
}
if("Respiratory Rate (count/min)" %in% names(cleaned)) {
cleaned <- dplyr::rename(cleaned,
resp_rate = `Respiratory Rate (count/min)`)
}
if("Active Calories (kcal)" %in% names(cleaned)) {
cleaned <- dplyr::rename(cleaned, active_cal = `Active Calories (kcal)`)
}
# Convert datetime variable to dttm type
cleaned$datetime <- lubridate::dmy_hm(cleaned$datetime,
tz = Sys.timezone())
return(cleaned)
},
create_util_data = function(start_date, end_date) {
# date range between start and end date
df <- tibble::data_frame(date = lubridate::ymd(as.Date(as.POSIXct(
seq(from = start_date, to = end_date, by = 1)))))
df$datetime <- lubridate::make_datetime(year = lubridate::year(df$date),
month = lubridate::month(df$date),
day = lubridate::day(df$date),
hour = 0L,
min = 0L,
sec = 0,
tz = Sys.timezone())
# MTWTFSS
df$day_of_week <- lubridate::wday(df$date, label = TRUE)
# weekend/weekday
weekend <- c('Sat', 'Sun')
df$day_type <- factor((df$day_of_week %in% weekend),
levels = c(FALSE, TRUE),
labels = c('weekday', 'weekend'))
# Month
df$month <- lubridate::month(df$date, label = TRUE)
return(tibble::as_data_frame(df))
},
get_fitbit_intraday = function(fitbit_user_email, fitbit_user_pw) {
cookie <- fitbitScraper::login(email = fitbit_user_email,
password = fitbit_user_pw)
start <- self$start_date
end <- self$end_date
char_start <- as.character(start)
char_end <- as.character(end)
# Build each intraday dataset: create list of lists, then rbind all
intraday <- list()
intraday$isteps <- NULL
intraday$idist <- NULL
intraday$ifloors <- NULL
intraday$iactive_min <- NULL
intraday$ical_burn <- NULL
intraday$ihr <- NULL
for (indiv_date in format(seq.Date(from = as.Date(start),
to = as.Date(end), by = "day"),
format = "%Y-%m-%d")){
char_date <- as.character(indiv_date)
intraday$isteps <- rbind(intraday$isteps,
fitbitScraper::get_intraday_data(
cookie, what = "steps", date = char_date))
intraday$idist <- rbind(intraday$idist,
fitbitScraper::get_intraday_data(
cookie, what = "distance", date = char_date))
intraday$ifloors <- rbind(intraday$ifloors,
fitbitScraper::get_intraday_data(
cookie, what = "floors", date = char_date))
intraday$iactive_min <- rbind(intraday$iactive_min,
fitbitScraper::get_intraday_data(
cookie, what = "active-minutes",
date = char_date))
intraday$ical_burn <- rbind(intraday$ical_burn,
fitbitScraper::get_intraday_data(
cookie, what = "calories-burned",
date = char_date))
intraday$ihr <- rbind(intraday$ihr,
fitbitScraper::get_intraday_data(
cookie, what = "heart-rate", date = char_date))
}
# Add weight (users can weigh many times/anytime in day)
intraday$weight <- fitbitScraper::get_weight_data(cookie,
start_date = char_start,
end_date = char_end)
# Join all, create date, time, datetime columns and drop dateTime
joined <- tibble::as_data_frame(Reduce(function(x, y) merge(x, y,
all = TRUE,
by = "time"),
intraday)
)
joined <- dplyr::select(joined, -dateTime)
joined$datetime <- lubridate::ymd_hms(joined$time, tz = Sys.timezone())
joined$date <- lubridate::ymd(as.Date(as.POSIXct(joined$datetime,
Sys.timezone())))
joined$time <-
lubridate::make_datetime(year = "1970", month = "01",
day = "01",
hour = lubridate::hour(joined$datetime),
min = lubridate::minute(joined$datetime),
sec = lubridate::second(joined$datetime))
joined <- plyr::rename(joined,
replace = c("active-minutes" = "activeMin"))
# create additional columns
joined <- dplyr::mutate(joined,
distanceKm = distance * MI_TO_KM,
weightKg = weight * MI_TO_KM)
# drop `calories-burned`, which is less informative than caloriesBurned
joined$`calories-burned` <- NULL
# reorder columns
joined <- dplyr::select(joined,
date,
time,
datetime,
steps, distance, distanceKm, floors, activeMin,
activityLevel, bpm, confidence, caloriesBurned,
defaultZone, customZone, weight, weightKg,
dplyr::everything())
return(joined)
},
# Returns a tibble of joined variables recorded daily
get_fitbit_daily = function(fitbit_user_email, fitbit_user_pw) {
cookie <- fitbitScraper::login(email = fitbit_user_email,
password = fitbit_user_pw)
start <- self$start_date
end <- self$end_date
char_start <- as.character(start)
char_end <- as.character(end)
daily <- list()
daily$steps <- fitbitScraper::get_daily_data(cookie, what = "steps",
start_date = char_start,
end_date = char_end)
daily$distance <- fitbitScraper::get_daily_data(cookie,
what = "distance",
start_date = char_start,
end_date = char_end)
daily$floors <- fitbitScraper::get_daily_data(cookie, what = "floors",
start_date = char_start,
end_date = char_end)
daily$minsVery <- fitbitScraper::get_daily_data(cookie,
what = "minutesVery",
start_date = char_start,
end_date = char_end)
daily$cal_ratio <- fitbitScraper::get_daily_data(cookie,
what = "caloriesBurnedVsIntake",
start_date = char_start,
end_date = char_end)
daily$rest_hr <- fitbitScraper::get_daily_data(cookie,
what = "getRestingHeartRateData",
start_date = as.character(start),
end_date = as.character(end))
daily$sleep <- fitbitScraper::get_sleep_data(cookie,
start_date = char_start,
end_date = char_end)[[2]]
# manipulate date, time, and datetime
daily$sleep$time <- as.POSIXct(daily$sleep$date)
joined <- tibble::as_data_frame(Reduce(function(x, y) merge(x, y,
all = TRUE,
by = "time"),
daily))
joined$date <- lubridate::ymd(as.Date(as.POSIXct(joined$time,
tz = Sys.timezone())))
joined <- dplyr::select(joined, -time)
joined$datetime <-
lubridate::make_datetime(year = lubridate::year(joined$date),
month = lubridate::month(joined$date),
day = lubridate::day(joined$date),
hour = 0L,
min = 0L,
sec = 0,
tz = Sys.timezone())
# create additional columns
joined <- dplyr::mutate(joined,
minRestlessAwake = joined$sleepDuration - joined$minAsleep,
sleepDurationHrs = sleepDuration / 60,
minAsleepHrs = minAsleep / 60,
restlessProp = (sleepDurationHrs - minAsleepHrs) /
sleepDurationHrs * 100,
distanceKm = distance * MI_TO_KM)
# rearrange columns:
joined <- dplyr::select(joined,
date,
datetime,
dateInForJavascriptLocalFormatting,
steps, distance, distanceKm, floors,
minutesVery, caloriesBurned, caloriesIntake,
restingHeartRate,
startTime, endTime,
startDateTime, endDateTime,
sleepDuration, sleepDurationHrs,
minAsleep, minAsleepHrs,
minRestlessAwake,
awakeCount,
restlessCount,
awakeDuration,
restlessDuration,
restlessProp,
dplyr::starts_with("sleepQuality"),
dplyr::starts_with("sleepBucket"),
clusters,
breaks,
dplyr::everything())
return(joined)
}
))
Try the lifelogr package in your browser
Any scripts or data that you put into this service are public.
lifelogr documentation built on May 2, 2019, 6:11 a.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
#' @include global_var.R
NULL
#' A \code{Person} object is a complete view of an individual over a certain
#' time period, as seen through data from multiple sources
#'
#' @description \code{Person} is an object that encapsulates an individual's
#' data over a specified date range (start and end date stored as \code{Date}
#' objects.
#' An individual consists of basic information, such as name, age,
#' and gender (a \code{list} with named elements), data from their self-tracking
#' devices such as Fitbit, Apple health, etc. (data from each source is a tibble
#' dataframe), individual goals such as target steps (\code{numeric}),
#' additional data from self-tracking apps or one's own collection system
#' (stored as a tibble dataframe), and ways of grouping the data a user may be
#' interested in, such as grouping by seasons, or comparing weekend to weekday
#' behavior and health (stored as a list of named dataframes, which each contain
#' group assignments).
#'
#' @docType class
#' @importFrom R6 R6Class
#' @importFrom tibble data_frame as_data_frame
#' @importFrom lubridate hour minute second make_datetime ymd wday month
#' @importFrom fitbitScraper login get_intraday_data get_weight_data
#' get_daily_data get_sleep_data
#' @importFrom dplyr select mutate
#' @importFrom plyr rename
#' @field fitbit_daily tibble dataframe of fitbit variables (for user account
#' info provided) observed daily.
#' Columns include:
#' \itemize{
#' \item{date: unique for each row (date)}
#' \item{datetime: includes date and time, time is an arbitrary
#' time, which is consistent for each day (date)}
#' \item{dateInForJavascriptLocalFormatting: chr}
#' \item{steps: total number of steps for that day (dbl)}
#' \item{distance: total distance for that day, in miles (dbl)}
#' \item{distanceKm: total distance for that day, in kilometers
#' (dbl)}
#' \item{floors: total number of floors for that day (dbl)}
#' \item{minutesVery: minutes 'very active' that day (dbl)}
#' \item{caloriesBurned: calories (kcal) burned that day (dbl)}
#' \item{caloriesIntake: calories (kcal) consumed that day, user
#' must input this, either into this data frame or into the fitbit
#' (dbl)}
#' \item{restingHeartRate: resting heart rate in beats per minute
#' (bpm) (dbl)}
#' \item{startTime: sleeping start time for that day (chr)}
#' \item{endTime: sleeping end time for that day (chr)}
#' \item{startDateTime: sleeping start date and time for that day
#' (chr)}
#' \item{endDateTime: sleeping end date and time for that day (chr)}
#' \item{sleepDuration: sleep duration for that day, in minutes
#' (int)}
#' \item{sleepDurationHrs: sleep duration for that day, in hours
#' (dbl)}
#' \item{minAsleep: time asleep that day, in minutes (int)}
#' \item{minAsleepHrs: time asleep that day, in hours, derived from
#' minAsleep (dbl)}
#' \item{minRestlessAwake: \eqn{sleepDuration - awakeCount} (int)}
#' \item{awakeCount: int}
#' \item{restlessCount: int}
#' \item{awakeDuration: int}
#' \item{restlessDuraton: int}
#' \item{restlessProp: proportion of sleep spent restless,
#' calculated as \eqn{restlessProp = (sleepDurationHrs -
#' minAsleepHrs) / sleepDurationHrs * 100} (dbl)}
#' \item{sleepQualityScoreB: dbl}
#' \item{sleepQualityScoreA: int}
#' \item{sleepQualityGraphicPercentA: dbl}
#' \item{sleepQualityGraphicPercentB: dbl}
#' \item{sleepBucketTextB: one of "ok", "good", "great" (chr)}
#' \item{sleepBucketTextA: one of "ok", "good", "great" (chr)}
#' \item{clusters: list of chr}
#' \item{breaks: list of chr}
#' }
#' @field fitbit_intraday tibble dataframe of fitbit variables (for user account
#' info provided) observed multiple times a day.
#' Columns include:
#' \itemize{
#' \item{date: unique for each row (date)}
#' \item{time: a combination of an arbitrary date ("1970-01-01")
#' and the time of the observation, generally in 5 minute intervals
#' (dttm)}
#' \item{datetime: includes date from `date` and time from `time`
#' (dttm)}
#' \item{steps: number of steps in 15 minute interval (dbl)}
#' \item{distance: distance traveled in 15 minute interval, in
#' miles (dbl)}
#' \item{distanceKm: distance traveled in 15 minute interval, in
#' kilometers (dbl)}
#' \item{floors: number of floors went up and down in 15 minute
#' interval (dbl)}
#' \item{activeMin: number of active minutes in 15 minute
#' interval (dbl)}
#' \item{activityLevel: hypothesized activity level, one of:
#' "SEDENTARY", "LIGHTLY_ACTIVE", "MODERATELY_ACTIVE", or
#' "VERY_ACTIVE" (chr)}
#' \item{bpm: average heart rate in 5 minute interval (int)}
#' \item{confidence: one of -1, 1, 2, or 3 (int)}
#' \item{caloriesBurned: calories (kcal) burned in 5 minute
#' interval (dbl)}
#' \item{defaultZone: chr}
#' \item{customZone: lgl}
#' \item{weight: weight, in lbs (dbl)}
#' \item{weightKg: weight, in kg (dbl)}
#' }
#' @field util tibble dataframe that maps each date in the date range to utility
#' information about that date
#' Columns include:
#' \itemize{
#' \item{date: unique for each row (date)}
#' \item{datetime: date from `date` and an arbitrary time (16:00:00)
#' (dttm)}
#' \item{day_of_week: day of the week, with Sun as first (ord)}
#' \item{day_type: weekend or weekday (fctr)}
#' \item{month: month, with Jan as first (ord)}
#' }
#' @field target_steps the person's target number of steps (numeric) for each
#' day (default 10,000)
#' @field start_date start of user's date range of interest (Date object)
#' @field end_date end of user's date range of interest (Date object)
#' @field user_info provided user info, such as "age", "gender", "name" (list)
#' @field groupings named list of dataframes, each with two columns - a known
#' variable, and group, with the group assignment for observations where
#' that variable has appropriate value
#' @field apple tibble dataframe of user's provided Apple Health data. These
#' columns depend on which columns are passed in by the user. However,
#' these columns match fitbit columns:
#' \itemize{
#' \item{datetime: dttm}
#' \item{steps: Original steps data for total number of steps in 60
#' minutes, but divided by 4 to match fitbit steps data, which is
#' the total number of steps in 15 minutes (dbl)}
#' \item{distance: Average distance in 15 minutes in miles.
#' Original distance data for total distance in 60 minutes,
#' but divided by 4 to match fitbit distance data, which is the total
#' distance in 15 minutes (dbl)}
#' \item{distanceKm: Average distance in 15 minutes in km.
#' Original distance data for total distance in 60 minutes,
#' but divided by 4 to match fitbit distance data, which is the total
#' distance in 15 minutes (dbl)}
#' \item{floors: Average number of floors in 15 minutes.
#' Original floors data for total number
#' of floors in 60 minutes, but divided by 4 to match fitbit floors
#' data, which is the total number of floors in 15 minutes(dbl)}
#' \item{bpm: average heart rate for the given hour, most users will
#' not have this data (dbl)}
#' }
#' @field addl_data dataframe of data from another source provided by user
#' @field addl_data2 dataframe of data from another source provided by user
#'
#' @section Methods:
#'
#' \describe{
#' \item{\code{Person$new(fitbit_user_email, fitbit_user_pw, user_info = NA,
#' apple_data_file, target_steps, addl_data,
#' addl_data2, group_assignments, start_date,
#' end_date)}}{Creates a new \code{Person}
#' with specified data, and data from provided Fitbit account. If
#' provided, start_date and end_date must be characters with "%Y-%m-%d".
#' All defaults are \code{NA} - user can provide sources of data
#' of interest.}}
#'
#' @examples
#' library("lifelogr")
#'
#' group_months <- data.frame("month" = c("Jan", "Feb", "Mar", "Apr", "May",
#' "Jun", "Jul", "Aug",
#' "Sep", "Oct", "Nov", "Dec"),
#' "group" = c(0, 0, 0, 1, 1, 1, 1, 1,
#' 1, 0, 0, 0))
#' ash <- Person$new(user_info = list("name" = "Ash", "age" = 26,
#' "gender" = "female"),
#' target_steps = 20000,
#' group_assignments = list("group_months" = group_months),
#' start_date = "2017-03-11",
#' end_date = "2017-03-12")
#'
#' \dontrun{
#' bailey <- Person$new(fitbit_user_email = "bailey@gmail.com",
#' fitbit_user_pw = "baileypw",
#' #apple_data_file = "apple.csv",
#' start_date = "2017-03-11",
#' end_date = "2017-03-12")}
#'
#' @export
#' @format An \code{\link{R6Class}} generator object
#'
Person <- R6::R6Class("Person",
public = list(
fitbit_daily = NULL, # dataframe of daily fitbit data
fitbit_intraday = NULL, # dataframe of intraday fitbit data
util = NULL, # util/admin data
target_steps = NULL,
start_date = NA, # optional start date of interest
end_date = NA, # optional end date of interest
user_info = NULL, # optional list of user info, such as "age", "gender", "name", etc.
groupings = NULL, # named list of dataframes each with 2 columns: 1. a known
# variable (date, weekend, etc.) and 2. "group" with the group assignment for
# observations with that variable of that value
apple = NULL, # apple data from csv
addl_data = NULL, # user's own df of other data
addl_data2 = NULL, # another df
initialize = function(fitbit_user_email = NA, fitbit_user_pw = NA,
user_info = NA,
apple_data_file = NA,
target_steps = 10000,
addl_data = NA, addl_data2 = NA,
group_assignments = NA,
start_date = NA, end_date = NA) {
self$addl_data <- addl_data
self$addl_data2 <- addl_data2
if (is.na(start_date)) {
start_date = "1990-01-01"
}
if (is.na(end_date)) {
end_date = Sys.Date()
}
self$start_date <- as.Date(strptime(start_date, format="%Y-%m-%d"))
self$end_date <- as.Date(strptime(end_date, format="%Y-%m-%d"))
self$target_steps <- target_steps
self$user_info <- user_info
self$util <- private$create_util_data(self$start_date, self$end_date)
if (!is.na(apple_data_file)){
self$apple <- private$load_apple_data(apple_data_file)
# user needs to pass in path from this directory to file
}
if (!is.na(fitbit_user_email) && !is.na(fitbit_user_pw)){
self$fitbit_intraday <- private$get_fitbit_intraday(fitbit_user_email,
fitbit_user_pw)
self$fitbit_daily <- private$get_fitbit_daily(fitbit_user_email,
fitbit_user_pw)
}
else {
self$fitbit_intraday <- NA
self$fitbit_daily <- NA
}
self$groupings <- group_assignments
}),
private = list(
load_apple_data = function(apple_data_file) {
raw_df <- readr::read_csv(apple_data_file)
cleaned <- tibble::as_tibble(raw_df)
# Remove Finish column because it's confusing
if(!is.null(cleaned$Finish)) {
cleaned$Finish <- NULL
}
# Rename columns to match Person$data_intraday and add additional columns
if("Start" %in% names(cleaned)) {
cleaned <- dplyr::rename(cleaned, datetime = Start)
}
# fitbit steps data is in 15 min intervals, while Apple steps data is in
# hour intervals
if("Steps (count)" %in% names(cleaned)) {
cleaned <- dplyr::rename(cleaned, steps = `Steps (count)`)
cleaned <- dplyr::mutate(cleaned,
steps = steps / 4)
}
# fitbit floors data is in 15 min intervals, while Apple floors data
# is in hour intervals
if("Flights Climbed (count)" %in% names(cleaned)) {
cleaned <- dplyr::rename(cleaned, floors = `Flights Climbed (count)`)
cleaned <- dplyr::mutate(cleaned, floors = floors / 4)
}
if("Heart Rate (count/min)" %in% names(cleaned)) {
cleaned <- dplyr::rename(cleaned, bpm = `Heart Rate (count/min)`)
}
# fitbit distance data is in 15 min intervals, while Apple distance data
# is in hour intervals
if("Distance (mi)" %in% names(cleaned)) {
cleaned <- dplyr::rename(cleaned, distance = `Distance (mi)`)
cleaned <- dplyr::mutate(cleaned,
distance = distance / 4,
distanceKm = distance * MI_TO_KM)
}
if("Respiratory Rate (count/min)" %in% names(cleaned)) {
cleaned <- dplyr::rename(cleaned,
resp_rate = `Respiratory Rate (count/min)`)
}
if("Active Calories (kcal)" %in% names(cleaned)) {
cleaned <- dplyr::rename(cleaned, active_cal = `Active Calories (kcal)`)
}
# Convert datetime variable to dttm type
cleaned$datetime <- lubridate::dmy_hm(cleaned$datetime,
tz = Sys.timezone())
return(cleaned)
},
create_util_data = function(start_date, end_date) {
# date range between start and end date
df <- tibble::data_frame(date = lubridate::ymd(as.Date(as.POSIXct(
seq(from = start_date, to = end_date, by = 1)))))
df$datetime <- lubridate::make_datetime(year = lubridate::year(df$date),
month = lubridate::month(df$date),
day = lubridate::day(df$date),
hour = 0L,
min = 0L,
sec = 0,
tz = Sys.timezone())
# MTWTFSS
df$day_of_week <- lubridate::wday(df$date, label = TRUE)
# weekend/weekday
weekend <- c('Sat', 'Sun')
df$day_type <- factor((df$day_of_week %in% weekend),
levels = c(FALSE, TRUE),
labels = c('weekday', 'weekend'))
# Month
df$month <- lubridate::month(df$date, label = TRUE)
return(tibble::as_data_frame(df))
},
get_fitbit_intraday = function(fitbit_user_email, fitbit_user_pw) {
cookie <- fitbitScraper::login(email = fitbit_user_email,
password = fitbit_user_pw)
start <- self$start_date
end <- self$end_date
char_start <- as.character(start)
char_end <- as.character(end)
# Build each intraday dataset: create list of lists, then rbind all
intraday <- list()
intraday$isteps <- NULL
intraday$idist <- NULL
intraday$ifloors <- NULL
intraday$iactive_min <- NULL
intraday$ical_burn <- NULL
intraday$ihr <- NULL
for (indiv_date in format(seq.Date(from = as.Date(start),
to = as.Date(end), by = "day"),
format = "%Y-%m-%d")){
char_date <- as.character(indiv_date)
intraday$isteps <- rbind(intraday$isteps,
fitbitScraper::get_intraday_data(
cookie, what = "steps", date = char_date))
intraday$idist <- rbind(intraday$idist,
fitbitScraper::get_intraday_data(
cookie, what = "distance", date = char_date))
intraday$ifloors <- rbind(intraday$ifloors,
fitbitScraper::get_intraday_data(
cookie, what = "floors", date = char_date))
intraday$iactive_min <- rbind(intraday$iactive_min,
fitbitScraper::get_intraday_data(
cookie, what = "active-minutes",
date = char_date))
intraday$ical_burn <- rbind(intraday$ical_burn,
fitbitScraper::get_intraday_data(
cookie, what = "calories-burned",
date = char_date))
intraday$ihr <- rbind(intraday$ihr,
fitbitScraper::get_intraday_data(
cookie, what = "heart-rate", date = char_date))
}
# Add weight (users can weigh many times/anytime in day)
intraday$weight <- fitbitScraper::get_weight_data(cookie,
start_date = char_start,
end_date = char_end)
# Join all, create date, time, datetime columns and drop dateTime
joined <- tibble::as_data_frame(Reduce(function(x, y) merge(x, y,
all = TRUE,
by = "time"),
intraday)
)
joined <- dplyr::select(joined, -dateTime)
joined$datetime <- lubridate::ymd_hms(joined$time, tz = Sys.timezone())
joined$date <- lubridate::ymd(as.Date(as.POSIXct(joined$datetime,
Sys.timezone())))
joined$time <-
lubridate::make_datetime(year = "1970", month = "01",
day = "01",
hour = lubridate::hour(joined$datetime),
min = lubridate::minute(joined$datetime),
sec = lubridate::second(joined$datetime))
joined <- plyr::rename(joined,
replace = c("active-minutes" = "activeMin"))
# create additional columns
joined <- dplyr::mutate(joined,
distanceKm = distance * MI_TO_KM,
weightKg = weight * MI_TO_KM)
# drop `calories-burned`, which is less informative than caloriesBurned
joined$`calories-burned` <- NULL
# reorder columns
joined <- dplyr::select(joined,
date,
time,
datetime,
steps, distance, distanceKm, floors, activeMin,
activityLevel, bpm, confidence, caloriesBurned,
defaultZone, customZone, weight, weightKg,
dplyr::everything())
return(joined)
},
# Returns a tibble of joined variables recorded daily
get_fitbit_daily = function(fitbit_user_email, fitbit_user_pw) {
cookie <- fitbitScraper::login(email = fitbit_user_email,
password = fitbit_user_pw)
start <- self$start_date
end <- self$end_date
char_start <- as.character(start)
char_end <- as.character(end)
daily <- list()
daily$steps <- fitbitScraper::get_daily_data(cookie, what = "steps",
start_date = char_start,
end_date = char_end)
daily$distance <- fitbitScraper::get_daily_data(cookie,
what = "distance",
start_date = char_start,
end_date = char_end)
daily$floors <- fitbitScraper::get_daily_data(cookie, what = "floors",
start_date = char_start,
end_date = char_end)
daily$minsVery <- fitbitScraper::get_daily_data(cookie,
what = "minutesVery",
start_date = char_start,
end_date = char_end)
daily$cal_ratio <- fitbitScraper::get_daily_data(cookie,
what = "caloriesBurnedVsIntake",
start_date = char_start,
end_date = char_end)
daily$rest_hr <- fitbitScraper::get_daily_data(cookie,
what = "getRestingHeartRateData",
start_date = as.character(start),
end_date = as.character(end))
daily$sleep <- fitbitScraper::get_sleep_data(cookie,
start_date = char_start,
end_date = char_end)[[2]]
# manipulate date, time, and datetime
daily$sleep$time <- as.POSIXct(daily$sleep$date)
joined <- tibble::as_data_frame(Reduce(function(x, y) merge(x, y,
all = TRUE,
by = "time"),
daily))
joined$date <- lubridate::ymd(as.Date(as.POSIXct(joined$time,
tz = Sys.timezone())))
joined <- dplyr::select(joined, -time)
joined$datetime <-
lubridate::make_datetime(year = lubridate::year(joined$date),
month = lubridate::month(joined$date),
day = lubridate::day(joined$date),
hour = 0L,
min = 0L,
sec = 0,
tz = Sys.timezone())
# create additional columns
joined <- dplyr::mutate(joined,
minRestlessAwake = joined$sleepDuration - joined$minAsleep,
sleepDurationHrs = sleepDuration / 60,
minAsleepHrs = minAsleep / 60,
restlessProp = (sleepDurationHrs - minAsleepHrs) /
sleepDurationHrs * 100,
distanceKm = distance * MI_TO_KM)
# rearrange columns:
joined <- dplyr::select(joined,
date,
datetime,
dateInForJavascriptLocalFormatting,
steps, distance, distanceKm, floors,
minutesVery, caloriesBurned, caloriesIntake,
restingHeartRate,
startTime, endTime,
startDateTime, endDateTime,
sleepDuration, sleepDurationHrs,
minAsleep, minAsleepHrs,
minRestlessAwake,
awakeCount,
restlessCount,
awakeDuration,
restlessDuration,
restlessProp,
dplyr::starts_with("sleepQuality"),
dplyr::starts_with("sleepBucket"),
clusters,
breaks,
dplyr::everything())
return(joined)
}
))
Try the lifelogr package in your browser
Any scripts or data that you put into this service are public.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.