Skip to content

Latest commit

 

History

History
86 lines (72 loc) · 3.55 KB

hrv-notes.md

File metadata and controls

86 lines (72 loc) · 3.55 KB
Raw

This document gives an overview on how HRV data is recoded by a Garmin device and how it can be used by ActivityLog2.

Recording HRV data

Garmin devices can record HRV data in the activity FIT files. This needs to be explicitly enabled on the device, but usually there is no option you can set on the device itself. ActivityLog2 allows creating a settings file that can be exported to the device and this can be used to enable HRV recording. Alternatively, there are various web sites describing how to do this for each device.

Reading HRV data

Since the device records the time interval between every heart beat, it can contain a lot of samples, so HRV data is not read during normal FIT file parsing and only summary HRV metrics are put into the database.

The actual HRV data can be read in a data-frame% by two functions (defined in "hrv.rkt"):

  • make-hrv-data-frame/file -- will read a FIT file
  • make-hrv-data-frame/db -- will read the FIT data stored in the database

Both functions will return #f if the FIT file does not contain HRV data. If you are using "al-interactive.rkt", the hrv-df provides a convenient shortcut. Evaluating the following program:

#lang racket
(require "al-interactive.rkt")

Will give an interactive prompt which allows loading, inspecting and saving HRV data into CSV files:

>
Opened C:\Users\alexh\AppData\Local\ActivityLog\Alex-AL.db
scratch.rkt> ;; HINT: Use "Activity/Copy session id to clipboard" menu
scratch.rkt> (define df (hrv->df 1717))
scratch.rkt> (df-describe df)
data-frame: 4 series, 6223 items
properties:
series:
              NAs           min           max          mean        stddev
  bpm           0            85           141        113.55         11.44
  delta-hrv     0             0         15527         34.75        272.65
  hrv           0           141         17040        607.45         348.7
  timestamp     0    1468032094    1468035492  1468033833.8       1107.83
scratch.rkt> (df->csv df "hrv.csv")
scratch.rkt>

The HRV data-frame% contains 4 series:

  • timestamp is a UNIX timestamp on when the recording was made. This series is ordered. Since there are several HRV values recorded every second, and the timestamp resolution is 1 second, there will be duplicates in this series.
  • hrv represents the time in milliseconds between heart beats, as recorded by the device. The series is stored as read from the FIT file and might contain bad values (e.g. missing heart beats are recorded as unusually large HRV values).
  • delta-hrv represents the difference between adjacent HRV values.
  • bpm is the HR data read from the FIT file. Can be used to correlate HRV values with the HR data recorded by the device.

The function compute-hrv-metrics in hrv.rkt can be used to compute some common HRV metrics, the results of this computation is stored in the database. The database stores metrics for the entire activity, but they can be computed for subsets. For example, to compute the metrics for the first 5 minutes of the activity:

scratch.rkt> ;; HINT: 1468032094 is the min value of "timestamp", see df-describe
scratch.rkt> (match-define (list start end)
               (send df get-index* "timestamp" 1468032094 (+ 1468032094 300)))
scratch.rkt> ;; HINT: see hrv-metrics struct definition (hrv.rkt) for the fields
scratch.rkt> (compute-hrv-metrics df #:start start #:end end)
(hrv-metrics
 63
 121
 69
 126
 0.12340842311459353
 298
 0.29187071498530853
 1014
 5209)
scratch.rkt>