search

aricooperdavis / Huawei-TCX-Converter

A makeshift python tool that generates TCX files from Huawei HiTrack files
MIT License
121 stars 23 forks source link

[ENHANCEMENT]pm-n / pm-b , Calories? #13

Open frarude opened 5 years ago

frarude commented 5 years ago

from Huawei Health Database Backup. Recorded with "Huawei Fit" device

Input tp=lbs;k=0;lat=52.276164;lon=10.550901;alt=109.90351867675781;t=1.560876178E12; (...) details see attached file tp=lbs;k=23;lat=52.275222;lon=10.551427;alt=103.27085876464844;t=1.560876223E12; tp=pm-n;k=1000022;v=43.051; tp=pm-b;k=1000020;v=39.051; tp=lbs;k=24;lat=52.275171;lon=10.55146;alt=104.21701049804688;t=1.560876225E12; (...) tp=lbs;k=31;lat=52.274825;lon=10.551657;alt=107.33544158935547;t=1.560876239E12; tp=s-r;k=1560876178563;v=164; tp=lbs;k=32;lat=52.274784;lon=10.551694;alt=107.3501205444336;t=1.560876241E12; (...) tp=lbs;k=39;lat=52.274493;lon=10.551933;alt=111.48688507080078;t=1.560876255E12; tp=pm-b;k=2000036;v=32.0; tp=lbs;k=40;lat=52.274454;lon=10.551967;alt=112.22886657714844;t=1.560876257E12; (...)

Output No idea what pm-b and pm-n (or b-p-m) are s-r is "steps-ran": k is the timestamp and v is number of steps (comes nearly every minute so it is almost the cadence in steps/minute but not exactly) p-m is "pace-median" k=distance in 1/100000 meters v=seconds since last p-m (or start)

Files Send me your HiTrack file for me to troubleshoot with. Screenshot_20190620-081318

HiTrack_1556813084888155681710664330001.zip Screenshot_20190620-081955 Screenshot_20190620-082019 Screenshot_20190620-082037

Anything else

frarude commented 5 years ago

I had a look at your example data. When I do a graph of p-m and one of b-p-m and scale them to the same length, then they are similar. Now, when I look at the end-values 338718117 for b-p-m and divide it by 544918117 for p-m I can see that this is (roughly) the factor of km to miles (0.62159 vs. 0,621371) So I would declare the b-p-m to "British-pace-median" ... Bildschirmfoto 2019-06-20 um 12 50 30

frarude commented 5 years ago

So now I believe to see the results of my final question:-) 👍 pm-n is "pace-median" per 100m (the sum of all v-values gives the approximate run-time missing the last distance less than 100m) and then probably: pm-b is something mile based (since the final sum of all v= values gives again the approx run time except the very last part)

Calories are then calculated of the distance and the runners weight (roughly by the formula 1kcal weight/kg distance/km ) That is in my example just 96kg x 8.46km = 812 kcal

aricooperdavis commented 5 years ago

Thanks so much for putting all this work in! I'd like to include this in a future update - I'm snowed under at the moment but will close this issue as soon as it's done :)