data-analysis ⊹ sports

Heart_rate_data_analysis1

Kazu Nozue

Tags: ggplot2; R; RStudio; Apple Watch 

library(XML)
library(tidyverse)
library(lubridate)
library(scales)
library(ggthemes)

Reading downloaded health data from an Apple watch.

path="/Volumes/data_personal/Kazu_blog/resources"
#zip <- paste(path, 'export_Mana_101320.zip', sep = '/')
zip <- paste(path, 'export_Mana_122620.zip', sep = '/')

unzip(zip, exdir = path)
Sys.sleep(3) # pause for 3 seconds to let your computer unzip it.

Parse the xml file.

xml <- xmlParse(paste0(path, '/apple_health_export/export.xml'))

summary(xml)
## $nameCounts
## 
##                           Record                    MetadataEntry 
##                           625705                            57389 
##      InstantaneousBeatsPerMinute                     WorkoutEvent 
##                             6004                              489 
##                  ActivitySummary HeartRateVariabilityMetadataList 
##                              354                              182 
##                          Workout                    FileReference 
##                              113                               47 
##                     WorkoutRoute                       ExportDate 
##                               47                                1 
##                       HealthData                               Me 
##                                1                                1 
## 
## $numNodes
## [1] 690333

Convert xml data into dataframes.

df_record <-   XML:::xmlAttrsToDataFrame(xml["//Record"])
df_activity <- XML:::xmlAttrsToDataFrame(xml["//ActivitySummary"])
df_workout <-  XML:::xmlAttrsToDataFrame(xml["//Workout"])
# df_clinical <- XML:::xmlAttrsToDataFrame(xml["//ClinicalRecord"])
# df_location <- XML:::xmlAttrsToDataFrame(xml["//Location"]) %>% 
#  mutate(latitude = as.numeric(as.character(latitude)),
#         longitude = as.numeric(as.character(longitude)))

Format the data

  • Note that time zone has to be specified (default is “UTC”, see details in help of ymd_hms (?ymd_hms). See also Time zone explanation.
df <- df_record %>%
  mutate(device = gsub(".*(name:)|,.*", "",device),
         value = as.numeric(as.character(value)),
         endDate = ymd_hms(endDate,tz="America/Tijuana"),
         date = date(endDate),
         year = year(endDate),
         month = month(endDate),
         day = day(endDate),
         yday = yday(endDate),
         wday = wday(endDate),
         hour = hour(endDate),
         minute = minute(endDate),
         type = str_remove(type, "HKQuantityTypeIdentifier")
         )
## Warning in mask$eval_all_mutate(dots[[i]]): NAs introduced by coercion
## Date in ISO8601 format; converting timezone from UTC to "America/Tijuana".

What type of health data we have?

df %>% select(type) %>% distinct()
##                                      type
## 1                                  Height
## 2                                BodyMass
## 3                               HeartRate
## 4                               StepCount
## 5                  DistanceWalkingRunning
## 6                       BasalEnergyBurned
## 7                      ActiveEnergyBurned
## 8                          FlightsClimbed
## 9                       AppleExerciseTime
## 10                       RestingHeartRate
## 11                WalkingHeartRateAverage
## 12                 HeadphoneAudioExposure
## 13                         AppleStandTime
## 14 HKCategoryTypeIdentifierAppleStandHour
## 15  HKCategoryTypeIdentifierMenstrualFlow
## 16 HKCategoryTypeIdentifierMindfulSession
## 17       HKCategoryTypeIdentifierBloating
## 18       HKCategoryTypeIdentifierDiarrhea
## 19               HeartRateVariabilitySDNN

Plot heartRate heatmap (weekly: max)

  # revised
  df %>%
  dplyr::filter(type == 'HeartRate') %>% 
  dplyr::filter(year==2020) %>%
  group_by(month,wday,hour) %>% 
  summarize(heartrate=max(value)) %>% 
  arrange(desc(heartrate)) %>%
  ggplot(aes(x=hour, y=wday,  fill=heartrate)) + 
    geom_tile(col = 'grey40') + 
    scale_fill_continuous(limit=c(0,210),labels = scales::comma, low = 'grey95', high = '#008FD5') +
    theme(panel.grid.major = element_blank()) +
    scale_x_continuous(
      breaks = c(0, 6, 12, 18),
      label = c("Midnight", "6 AM", "Midday", "6 PM")
    ) +
    scale_y_reverse(
      breaks = c(1,2,3,4,5,6,7),
      label = c("Sunday", "Monday", "Tuesday", "Wednesday", "Thursday", "Friday", "Saturday")
    ) + 
    labs(title = "Weekly max HeartRate Heatmap",
         caption = '@taraskaduk | taraskaduk.com',x="Time",y="")  + 
    coord_equal() + facet_wrap(.~month,ncol=3)
## `summarise()` regrouping output by 'month', 'wday' (override with `.groups` argument)

Monthly Mean Heart rate , everyday calendar

df3 <- df %>%
  dplyr::filter(type == 'HeartRate') %>% 
  dplyr::filter(year==2020) %>%
  group_by(date,hour,minute) %>%
  summarize(heartrate=mean(value)) %>%
  group_by(date,hour) %>% #View()
  summarize(mean.heartrate=mean(heartrate)) %>%
    arrange(desc(mean.heartrate)) %>%
    mutate(yday=yday(date),
           wday=wday(date),
           mday=mday(date),
           month=month(date),
           day=day(date)) %>%
    group_by(date,month,mday,hour) %>% 
  arrange(desc(mean.heartrate))
## `summarise()` regrouping output by 'date', 'hour' (override with `.groups` argument)
## `summarise()` regrouping output by 'date' (override with `.groups` argument)
  pHRmonth <- ggplot(df3,aes(x=hour, y=mday,  fill=mean.heartrate)) + 
    geom_tile(col = 'grey40') + 
    scale_fill_continuous(labels = scales::comma, low = 'grey95', high = '#008FD5') +
    theme(panel.grid.major = element_blank(),axis.text.x=element_text(angle=90)) +
    scale_x_continuous(
      breaks = c(0, 6, 12, 18),
      label = c("Midnight", "6 AM", "Midday", "6 PM")
    ) +
    scale_y_reverse(
      breaks = 1:31,label=c(1," ",3," ",5," ",7," ",9," ",11," ",13," ",15," ",17," ",19," ",21," ",23," ",25," ",27," ",29," ",31)
    #  label = c("Sunday", "Monday", "Tuesday", "Wednesday", "Thursday", "Friday", "Saturday")
) + 
    labs(title = "Monthly mean heartrate Heatmap",
#         caption = '@taraskaduk | taraskaduk.com',
x="Time",y="")  + 
    coord_equal() + facet_wrap(.~month,ncol=3)
  pHRmonth

Make notes on graph by geom_text()

  • To writes texts on specific place in a graph, use separate data for the text info.
# data frame for text
df3.text <- tibble(month=c(3,5,9,10),mday=c(19,20,15,6),hour=12,mean.heartrate=100,label=c("Shelter in place","Synchro practice restart","Synchro fall training starts","Synchro team practice starts"))
pHRmonth + geom_text(data = df3.text,mapping=aes(label = label),size=3)