I think I found the problem.
In lines 41 and 54 the script uses start_time and end_time as index in the stream val. This works, if time is without gaps and the time increment is 1 sec. In my case there are gaps (probably through autostop), leading to a difference in index and time. If I change them to start_index and end_index it works perfectly.
1 Like
Nice! I will change the script on te shared version. Thanks for troubleshooting!
I added Torque (Nm) analysis to the list; @MedTechCD you may want to double check that it’s correct. 
Great diagram, thank you! Could you make it for W/kg, too? It would be great to discuss group-rides with group members: the individual watts are less relevant in that context, but rather the “normalized effort” (on a rolling terrain), which is the W/kg