Runner's Dash

• Goals
• Implemented visualizations
• Installation
• Usage
  • 1. Extraction
    • More details
  • 2. Dashboard
• A note on data privacy

Project goals

• Create an interactive dashboard to visualize running data alongside other health metrics, and be able to do some exploratory data analysis with personal health data.
  • Extract health data from exported Apple Health (XML) file and store them on a local database.
  • Use Plotly Dash to generate a dashboard.

Implemented visualizations

1. Time series plots across one year, binned weekly and monthly.

    - Total Distance (sum)
    - Total Duration (sum)
    - Total Energy Burned (mean)
    - Total Elevation Ascended (mean)
    - Avg. Pace
    - Avg. VO2Max (if available)
    - Avg. Weight (labeled as 'BodyMass')
    - Avg. METs
    - Avg. Resting Heart Rate
    - Avg. Heart Rate Variability SDNN
    - Avg. Respiratory Rate
    - Avg. Blood Pressure Diastolic
    - Avg. Blood Pressure Systolic
    - Avg. Menstrual Flow

2. Total distance run per day across one week.
3. Histograms (number of runs across a year/years) and a line plot of their associated avg. distance per run.

    - Day of the Week
    - Hour of run start time
    - Total run distance
    - Daily total distance, duration, avg METs

4. Various running/health metrics against temperature and humidity (for days with outdoor runs).

Installation

1. Clone this repository under your chosen directory

git clone https://github.com/nadinetab/runnersdash.git

2. Navigate to the project directory and install dependencies

pip install -r requirements.txt

Usage

1. Extract, store, and clean Apple Health data

Relevant file(s): exporthealthdata.py

1. Under the Health app settings (click on your profile picture), there will be a link to 'Export All Health Data'. This will download the file export.zip to your device.

2. Move this file to the same directory you cloned this repository to and open it to extract. This will create a new folder called apple_health_export where all your health data reside.

3. To run exporthealthdata.py:

   • If you want to extract all entries within export.xml:

         $ python exporthealthdata.py
     

   • If you already have a database (.db) file of a previous export and just need to add new entries since the last export:

         $ python exporthealthdata.py --append
     

     assuming there is an existing db file within the data/ subdirectory.

In detail, usage of exporthealthdata.py:

```
python exporthealthdata.py [-o --open-file </path/to/export.xml>]
                             [-a --append] [-v --version]
                             [-w --workouts] [-r --records]

OPTIONAL ARGUMENTS:

-o --open-export </path/to/export.xml> : 
    Determine export.xml file to read.

-a --append (bool) : 
    If passed in, script will find the latest version of a db file within the 'data/' subdirectory and append to the database any data with 'startDate' >= latest export date of the .db file. For use if you have previous exports already stored in a database and just need to add new data entries since your last export.
    Default is --no-append.

-v --version (bool) : 
    If true, the version number of extractapplehealth.py is appended to the name of the resulting database .db file.
    Default is --no-version.

-w --workouts <list of Workout names> : 
    List (space-separated) of table names (as it appears) in the database file of type Workout to process.

    Default: Running

    Note: Tables that are not listed in the default list have not yet been implemented within the data processing routines.

    Example Workout tables:
        - Running
        - Barre
        - Flexibility
        - Walking
        - Other
        - Skiing

-r --records <list of Record names> :
    List (space-separated) of table names of type Record to process.

    Default: 
        MenstrualFlow RestingHeartRate VO2Max BodyMass HeartRateVariabilitySDNN StepCount RespiratoryRate BloodPressureDiastolic BloodPressureSystolic

    Note: Tables that are not listed in the default list have not yet been implemented within the data processing routines.

    Example Record tables:
        - HeartRateVariabilitySDNN
        - VO2Max
        - StepCount
```

Outputs:

This step will produce the following things within data/ subdirectory:

1. A database (.db file) containing data from export.xml organized and stored into tables.
2. A set of five CSV files, containing daily/weekly/monthly aggregates as well as resampled running data, given default arguments for --workouts and --records parameters.

2. Launch dashboard

Related file(s): app.py

Launch the Dash app with the following:

python app.py

A note regarding data privacy

All health data is stored locally (and remains local).

Future implementations?

1. 'Sandbox' page with customizeable and extensive graph options and all available metrics to plot.
2. Compatibility with Google Fit and Samsung Health datasets?
3. A heatmap-view page, with calendar days on the y-axis and months of the year on the x-axis. Each cell represents that day's data. Clicking on the cell will produce visualization of that day's (assuming there is only one run/day) route and other run details (temp, humidity, plots of heart rate during run, etc.).