Inconsistent Recovery Point formulas

[Poncho]

Hi Ale

https://github.com/GoldenCheetah/GoldenCheetah/blob/4dd3c12a2c27153eef11a737bdf7c028a43c42af/src/FileIO/FilterHRV.cpp#L307

and

https://github.com/GoldenCheetah/GoldenCheetah/blame/4dd3c12a2c27153eef11a737bdf7c028a43c42af/src/Metrics/HrvMetrics.cpp#L830

use two different formulas for hrv recovery points.

It used to be the same formula until
https://github.com/GoldenCheetah/GoldenCheetah/commit/76dedfb4822db117c971f15a05e98e83f6d7fc1a

Might be on purpose, might me a bug. Just letting you know.

Best regards,
Poncho

[Ale Martinez]

It is a bug, long story short:

- Initially I tried to mimic HRV4Training Recovery Points so I did a linear regression on my own data and came up with  1.5*log(RMSSD)+2

- later I read a Marco Altini's (HRV4Training author) post saying it was just 2*log(RMSSD) so I changed it, but not in all places

I will fix it.

More recently Marco said to me if he had to start again from scratch it would just use RMSSD directly and It seems the tendency now.

Thank you for the heads up.

Ale.