HMs DIY #1-6

Hi

Thanks for the link. So, there would be 2 values, one np for the last 30s and one overall np for the whole session, right?

The "GoldenCheetah" calculates one value for the whole session afterwards, so no live calculation.

If you want the np for the whole session I would have to store every value for every second, I would probably run out of memory pretty soon. The bigger problem I see, is that the time to calculate the value would increase for every second. It would have to do a loop 7200 times after 2 hours.

Which one should be displayed?

EDIT: As I think about it, I think that none of the above seems to be a problem, I would just need a array of 30 values and a total value... hmm.. let me brain *** this a little bit more

Hi I've been trying to get my head round it too. My maths may be a little suspect. Hope I haven't set too much of a puzzle.

I was thinking that maybe each 30 second avg power sample could be added to the average of the previous samples, divide the result by 2 to get the overall average 30s power. Then perform the NP calculation on the result.?

Ie

Sample 1: 50 w

Calc NP based on 50w

Sample 2: 60 w

Sample 1 + Sample 2 = 110w / 2 = 55w avg pwr

Calc NP based on 55w

Sample 3: 45 w

Avg of Sample 1 & 2 + Sample 3 = 100w / 2 = 50w avg pwr

Calc NP based on 50w

And so on.... Would that work?

Hi

That would not work, because you would treat the sum of the old values the same as one new value.

But I think I know how to implement it. It needs a array of 30 values, a total value and a value for the number of seconds.

Ahh I understand. I did say my maths could be a bit suspect I'm glad someone knows what they're doing

What else came into my mind:

on an Edge you can declare if 0 Watt (rolling down a hill) should be in- or excluded from average-calculation. I always set power and also cadence zero-readings to „exclude“, because I‘m interested in statistics for the time I‘m working.

What does that mean for NP...?

For the native garmin datafield calculations i would imagine it would take the 30s avg excluding the zeros & calculate the NP on that value.

I guess that a CIQ app would be reading in the raw power data so unless the dev chose to exclude zeros they would be included in the NP calculation. Resulting in a lower NP.

I have mine set to include zeros as I prefer to see the overall power stats as if I'm not working my body is perhaps recovering so I was thinking it would skew the figures

I just checked the stats of some of my rides and it seems NP always includes zero-values.

New version 1.2.6 is out!

Excellent turnaround, Thank you. Will test over the weekend