FAQ: Filling Holes in performance_condition, enhanced_respiration_rate, potential, and stamina

A place for the community to help each other out with getting the most out of the Combine FIT, GPX or TCX files for Strava Upload Tool.
Post Reply
User avatar
fulmar2
Site Admin
Posts: 255
Joined: Wed Nov 25, 2020 4:21 am
Contact:

FAQ: Filling Holes in performance_condition, enhanced_respiration_rate, potential, and stamina

Post by fulmar2 »

When GOTOES merges your files, it attempts to preserve as much of the original data as possible. One thing I noticed recently was that for the ConnectIQ data, people sometimes merged two overlapping activities where one file was using "1 second recording"... and the other file was using "smart recording". The problem happened when the ConnectIQ data was recorded in the "smart recording" file... and the remaining data was recorded in the "1 second recording" file. Because GOTOES tries to remain as true as possible to the original data AND allow the user to grab as much data as possible from their available sources, we were getting "holes" in the data. For any given data point in the "1 second recording" file, there may (or may NOT) be a corresponding value for the "performance_condition, enhanced_respiration_rate, potential, and stamina" values.

The resulting merge looks like this on Garmin Connect:

Screen Shot 2024-11-16 at 6.05.59 PM.png

That isn't very good, as you can see, the stamina dropped to zero a few times. I already had a "back fill" feature put in place to address the issue where sometimes a GPS device would not immediately start recording performance_condition, enhanced_respiration_rate, potential, or stamina in the early moments of a file... so I decided to take that same code and also run a "forward fill." The decision is that if any of the values are empty for these fields, GOTOES will smooth out the empties and fill the values with the last known "good" value. After the code update, you can see that the graph looks much better on Garmin Connect.


Screen Shot 2024-11-16 at 6.05.32 PM.png

Hope that work-around helps some of you out!!
Post Reply