Upper Trimp & Tss numbers do not match lower ones

[gringojack]

trimptss.jpg (97.42 KiB) Viewed 12803 times

[/attachment]
Hi

As shown on the attached image, upper trimp and tss numbers do not match those in the training analysis plugin pane.
Sometimes they do!... No clue to explain

[old_man_biking]

The value on the Training Analysis detail page are dynamically calculated by Trainint Analysis plugin.
The values in the list above seem to be custom data fields. TA plugin does NOT fill these in - I don't know for sure which plugin does so.

Which values are "correct"? Hover the mouse over the "TRIMP" row in the Training Analysis plugin detail page (your 94.7 value) and the plugin will show you the calculation formula.
Which corresponds to the one found in section 5 of http://fellrnr.com/wiki/TRIMP or here: http://www.fmh.utl.pt/agon/cpfmh/docs/d ... Fadiga.pdf .
TSS calculation follows the procedure described here: http://lists.topica.com/lists/wattage/r ... =907028398

Btw, a major factor when calculating TSS is your threshold power. TA plugin uses the value shown below on its page (170W in your case). Whichever other plugin has calculated the value in the custom data field will have used a different value.
Btw 2, 170W is a VERY LOW value for threshold power, much too low if you're a slightly well-trained normal-sized male and not a couch potato.

Cheers,
OMB

[texmurphy]

Other plugin which computes Trimp / TSS is Training Load.

[gringojack]

thank you Tex

[tndmbkr]

..your TRIMP shows 259.5 and in mechgt's trainingload plugin instead I get 244.....

....forgot about a great feature of TA plugin....Have a look at Other Tasks : Select view: TA plugin : Activity Category Extensions.This page lets you define several ST properties per activity type.....

Other plugin which computes Trimp / TSS is Training Load.

The value on the Training Analysis detail page are dynamically calculated by Trainint Analysis plugin...... Hover the mouse over the "TRIMP" row in the Training Analysis plugin detail page (your 94.7 value) and the plugin will show you the calculation formula......

select "Automatic mode in TL" then do what Dirk says....

Just a few quotes from the past months.
But they do not answer my question why TRIMP as calculated by TA from OMB differs from TRIMP as calculated by TrainingLoad from Mechgt.

In my case the difference is always huge (200 from TA, 300 from TL)
The difference in TSS is negligible.
Other values for Data Smoothing do not change this.

========================
Formula as used by OMB:
TRIMP = duration * %HRReserve * 0.64 * exp (%HRReserve * 1.92)
where %HRReserve=(HRavg-HRrest)/(HRmax-HRrest)

Mechgt in automatic mode:
Automatic Mode uses Max HR and Resting HR (set previously) to calculate TRIMP.
Set your max & rest HR in athlete settings of ST.

Mechgt in manual mode:
Create a new heart rate category with approximately 10 heart rate zones evenly spread between a low HR and your max hr. Each zone might cover be 6 - 8 bpm, and your lowest zone would be relatively inactive.

Imagine an activity of 1 hr duration with HR=130 from start to end
My HRrest=65, and HRmax=165
TA calculates TRIMP=87. This according to the formula
Mechgt in automatic mode calculates TRIMP=97

Maybe it is important to observe that TA asked for a datatrack for this imaginary activity
TL was satisfied with a mention of the average HR
=================================
Text added after OMB's reply below:
I made a datatrack of this imaginary activity which starts with 100 bpm, and ends with 160 bpm.
HRavg still equals 130 bpm
One should expect higher values for TRIMP in this case. BUT:
TA calculates TRIMP=87 (again). This is not according to the formula. The formula calculates TRIMP=100
TL in automatic mode calculates TRIMP=108

Then I made an irregular datatrack of this imaginary activity with HRvalues varying from 100 to 160.
HRavg still equals 130 bpm
TA calculates TRIMP=87 (again).This is not according to the formula. The formula calculates TRIMP=126
TL in automatic mode calculates TRIMP=136

[old_man_biking]

TA plugin transparently tells you how it calculates TRIMP. Your example shows that it does so correctly. I've taken this formula from renowned resources.
I can't speak for mechgt's plugins. I don't know how they do the calculation.
Your observation that TA plugin requires a hr track to be present, instead of accepting the entered "avg. hr" value is correct. I'll change that so that it uses the entered avg. values if no track is present.

Cheers,
OMB

[texmurphy]

I wonder about possible different treatment of paused time(s) during the workouts.

[Burnt Toast]

tnd

Do you have your ST HR Zones and Power Zones set to the same values as OMB's Training Analysis plugin?

I use both TL and TA. They both display the the same values within a few tenths of each other.

[Burnt Toast]

TA plugin transparently tells you how it calculates TRIMP. Your example shows that it does so correctly. I've taken this formula from renowned resources.
I can't speak for mechgt's plugins. I don't know how they do the calculation.
Your observation that TA plugin requires a hr track to be present, instead of accepting the entered "avg. hr" value is correct. I'll change that so that it uses the entered avg. values if no track is present.

Cheers,
OMB

On a side note, maybe not related but has to do with TA, I may have discovered a hiccup but I'm not quite sure where its coming from.

It seems that in TA, the calculations are changing after I compress the data using ST's compression plugin. Yesterdays workout was uploaded to ST.mobi and synced this morning to ST3. After ST opens and I looked at TA, my data looked like this:

TSS = 118.1 (matched TL)
IF = 0.97
NP = 258
VI = 1.09

After I compressed my data and saved it. I went back to TA and found my data to be:

TSS = 122.6
IF = 0.99
NP = 263
VI = 1.11

[tndmbkr]

Do you have your ST HR Zones and Power Zones set to the same values as OMB's Training Analysis plugin? I use both TL and TA. They both display the the same values within a few tenths of each other.

In order to be sure, I installed a clean version of ST (+ GFD, TA, , AfterImport, TL) in a clean W7 install.
According to OMB's manual I entered weight, HRmax, HRrest, gender in the Athlete view for date 01-01-2014
I created a standard Heart Rate Category for TA (Select View Categories, TA, %HRR). This new category matches my HRrest and HRmax
According to Mechgt's manual I assigned this Standard Heart Rate Categorie to TL
Settings of ST: include stopped in time & distance totals; watch pauses not included
HR Data smooting = 6sec

I imported an activity (december 2014) from my Edge with GFD, and AfterImport.
The activity was straightforward without traffic jams, flat tyres, etc; watch time= moving time= total time
I had put some gel on the HRstrap for more accuracy.
I used a Garmin heartstrap and a Garmin Edge500.
The Edge recorded my heartbeat every second (no smartrecording)

TA calculates TRIMP=242
TL auto calculates TRIMP=274
TL manual calculates TRIMP=275

In this case the difference between TA and TL is only 32 points.
More often the difference is much larger.
When I recalculate the TRIMP of this activity in a spreadsheet, the outcome is inbetween TA and TL, but much closer to TA. (to perform this calculation I had to change more than one out of bound value manually, I also noticed many zero heartbeats)

Then I created three different datatracks with HRavg=130 (duration 1 hr)
One with HR=130 constant, one with HR rising from 100 to 160, one with three intervals HR=160/100
TA calculated for all three tracks TRIMP=87. Which cannot be true.
When I recalculated TRIMP, according to OMB's formula, in a spreadsheet , I got these outcomes: 87, 100, 126
TL calculates TRIMP= 97, 108, 136