About

This repository contains R scripts for connecting to the Sports Data Valley (SDV) API to retrieve wearable data from users (Fitbit, Garmin) in research groups, quantify missing or incomplete heart rate time series, and provide interactive visualization tools for researchers.

The pipeline is part of the Stress in Action consortium (Research Theme 2), which focuses on reliable real-time measurement of daily life stress and its impact on health using wearable technology.

How to Use This Repository

Step 1: Setup and Authentication

Create a .Renviron file with your API credentials and group ID:

BASE_URL=https://app.sportdatavalley.nl
API_PATH=/api/v1
AUTH_HEADER=Bearer your-token
GROUP_NO=your-group-id

Install required R packages:

install.packages(c("httr", "lubridate", "rjson", "here", "tidyverse", "plotly", "data.table"))

Step 2: Retrieve Group Information

Use the provided function get_group_info() to list all members in your SDV group. This step creates a profile_list object with user IDs and names.

Step 3: Download Activity Data

Fetch all activity data for your group using the get_all_group_rec_act() function. This retrieves recent VDOs (Versioned Data Objects) and parses structured data for:

• User ID and name
• Data type (Fitbit or Garmin)
• Date of measurement (dateTime or calendar_date)
• Heart rate time series (heart_intra_day for Fitbit or time_offset_heart_rate_samples for Garmin)
• Versioned Data Object ID

Step 4: Identify Missing or Incomplete Data

The script checks whether:

• The time series is NULL
• The length of the series is less than 50% of expected values:
  • Fitbit: (60 × 24) / 2 samples
  • Garmin: (4 × 60 × 24) / 2 samples

Filtered data is saved in filtered_data, grouped by user ID and including:

• User Info: first name, last name
• Record Details: date, time series length, dataset ID

Step 5: Visualize Missing Data

The visualize_missing_data.R script creates an interactive Plotly scatter plot:

• X-axis: Date of measurement
• Y-axis: Time series length
• Dropdown menu: Select a user to filter the view
• Hover details: User name, ID, date, samples, dataset ID

Step 6: Analyze Individual Datasets

Use get_data() to retrieve and explore specific VDOs:

• Extract heart rate time series
• Convert to a tidy data frame
• Generate an interactive line plot of heart rate values over time using Plotly

Example Visualization

An example output is an interactive scatter plot where each point represents a daily record for a user. Researchers can quickly identify:

• Days with missing time series (0 samples)
• Days with incomplete coverage

(Add screenshot here after running the pipeline)

API Documentation and Resources

• Sports Data Valley API Docs
• Stress in Action Project