Running features for v3.3 and 4.0 and VDOT Calculator

[Ale Martinez]

Now 3.2 is in the wild and the users have had some time to test the running features I think it would be a good time to think about the next versions, anyone who wish to contribute his ideas is welcome to comment here or to make a feature request at the tracker.

For my part I recently added an initial implementation of a little VDOT Calculator to master here is how it looks in metric and imperial units:

Even when the CP and W' Estimator in v3.2 can be used for running requires 2 relatively close tests and its use is not that common, Daniels' VDOT on the other hand can be used to estimate training paces (including T-Pace) using one race result and I find it very useful.

[art...@gmail.com]

Hi Ale,

good news. How do you implement VDOT? Do we find a new entry in Pace Zones?

cheers
Andy

[Ale Martinez]

Hi Andy, it hangs of the tools menu, near CP and W' Calculator.

Interestingly Daniels' paces are prescriptive, but if you use the T-Pace calculated from VDOT as your CV in Pace Zones, they align nicely with the descriptive Skiba's zones GC uses by default:

E-Pace is at the border between Z1-Z2

M-Pace in Z3

T-Pace near the middle of Z4

I-Pace in Z5

R-Paces in Z6

Cheers, Ale.

[bardfr...@gmail.com]

Would it be possible to add the ability to import running activities from a Suunto Ambit file format (.sml ) ??

[bardfr...@gmail.com]

Oh also...could it be possible to be able to assign gear/shoes...to any given running activity ?

[Ale Martinez]

This is possible in the current version, just define a gear or shoes data field with the shoes you want to track as values, in the same way as the Equipment field in the wiki example: https://github.com/GoldenCheetah/GoldenCheetah/wiki/UG_Preferences_Data%20Fields, then you can filter/search any chart by shoe, as explained here: https://github.com/GoldenCheetah/GoldenCheetah/wiki/UG_Special-Topics_SearchFilter

[bardfr...@gmail.com]

Oh nice, thanks!

Then, is it possible to see how many miles were run on one particular shoe then ?

[Ale Martinez]

El jueves, 17 de septiembre de 2015, 18:05:00 (UTC-3), bardfr...@gmail.com escribió:

Oh nice, thanks!

Then, is it possible to see how many miles were run on one particular shoe then ?

Yep, once you define the field and populate it in Details just filter/search Trends>Summary for an appropriate date range, see below:

 

[bardfr...@gmail.com]

Perfect! I should have made more research before asking that, sorry!

Now, what about the Suunto sml file import ? is this one already available ? if Yes, Im jumping on GC and transferring all my Sporttracks data on it !!

[Ale Martinez]

El jueves, 17 de septiembre de 2015, 18:29:17 (UTC-3), bardfr...@gmail.com escribió:

Now, what about the Suunto sml file import ? is this one already available ? if Yes, Im jumping on GC and transferring all my Sporttracks data on it !!

I'm afraid not, but Sporttracks fitlog format is supported so you could try to export your data from Sporttracks in that format to import in GoldenCheetah.

If you want to open a feature request for native sml import please do it at the tracker: https://github.com/GoldenCheetah/GoldenCheetah/issues

[bardfr...@gmail.com]

Good to know thanks!

Ive added an "issue" to add the SML suunto import

In the meantime, I will look at using the fitlog format to transfer my data from SportTracks to GoldenCheetah.

Thanks,

Frederic

[Mark Liversedge]

FITLOG is a terrible format. Please be careful. They remove data series like speed and distance and derive from GPS instead (!!!!).

Mark

[Jan de Visser]

There is a TCX export plugin in the SportTracks plugin catalog. I seem to remember it's free to use for a month or so after which you need to pay (but ot may also be free. Can't really remember).

 

I don't think there's plugins for more sane formats (.fit), but TCX is not all evil I think.

 

jan

--
_______________________________________________
Golden-Cheetah-Users mailing list
golden-che...@googlegroups.com
http://groups.google.com/group/golden-cheetah-users?hl=en
---
You received this message because you are subscribed to the Google Groups "golden-cheetah-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to golden-cheetah-u...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[bardfr...@gmail.com]

Thank you for the recommendations, indeed, I didnot know that the TCX format was not the best.

Do you think then that it will be possible one day to import from a Suunto sml file ? or will I have to convert from sml to tcx every time I want to import an activity into GC ?

To unsubscribe from this group and stop receiving emails from it, send an email to golden-cheetah-users+unsub...@googlegroups.com.

[Ale Martinez]

Can you post some sml files? If you have for different sports one for each would be helpful

[bardfr...@gmail.com]

Please find attached both running files. I would think that my cycling activities are of the same type as I don't have any accessories except my heart rate belt ?

Thank you

Frederic

[Ale Martinez]

I can't find any laps in these files, is that right?

I ran this simple python script to convert them to .gpx and they look pretty good when imported to GC, can you verify them?

[bardfr...@gmail.com]

Correct, there was no lap in  my files. Attached is a file with 3 laps.

I impprted the gpx files into my GC 3.1 and it seems to work fine! I cant see the Heart rate or temperature but i think its because im still a newbie at GC.
Frederic

[Ale Martinez]

HR and Temperature are present in the gpx and imported to GC, BTW I would recommend to upgrade to 3.2:

WRT to laps they are not supported by the Python script, I found these two but I'm not sure how to interpret them:

<Sample>

<UTC>2015-06-10T12:27:35.994Z</UTC>

<Time>2907.003</Time>

<Events>

<Lap>

<Type>Manual</Type>

<Duration>2757.4</Duration>

<Distance>7455</Distance>

</Lap>

</Events>

</Sample>

<Sample>

<UTC>2015-06-10T12:44:31.507Z</UTC>

<Time>3922.516</Time>

<Events>

<Lap>

<Type>Manual</Type>

<Duration>1015.4</Duration>

<Distance>2446</Distance>

</Lap>

</Events>

</Sample>

How do they look on Movescount or Sporttracks?

[Serban Mestecaneanu]

Hi Ale,

Here are some sml from me, the running one has laps as well. The files are from run, crosstrainer and bike. The nice part with the sml files is that include HRV data, plus some other specifics from Suunto, like Peak training effect, battery level, barometric pressure etc. I will have to post in separate posts due to size restrictions.

All the best,

Serban

[Serban Mestecaneanu]

And here is the bike

Cheers,

Serban

[Ale Martinez]

Thanks Serban, the format looks interesting, periodic samples are separated from GPS coordinates samples, I will try to adapt the GPX reader in GC and see how it works.

[bardfr...@gmail.com]

Thanks Serban for all this additional info. Indeed, the Suunto SML format encapsulates a lot of useful info (does GC have the ability to analyze HRV data ? ST doesn't) and that would be great that GC could read and display all that.

Ale, I am reluctant to install Debian unstable just to have GC 3.2. I prefer waiting that the 3.2 is in the stable repository.

Here is what the laps look like in Movescount, I hope it helps you, I am not sure that I understood your request, sorry

Also, what is WRT ? what does it mean ? 

Frederic

[Ale Martinez]

Based on TCX/GPX support in GC I implemented an initial version of SML import, people compiling from source can apply this commit to try it, included are the test files contributed in this thread.

Date/Time and Sport are taken from the header and other information from Samples, periodic ones to generate records with GC supported data and GPS ones for coordinates.

Data not currently supported by GC, s.t. barometric pressure or HRV data is ignored.

Includes lap and indoor activities support.

Would be interesting to have some bike file with power to test, it should work as is, and lap swimming files, probably it will require modifications.

[bardfr...@gmail.com]

WOW! That's what I call some SERIOUS support from GC !!! Thanks a bunch Ale! (We're far from the poor support offered from the unfortunate dying ST, another good reason for me to migrate over GC!)

As I am still discovering GC, I am working on setting up my environment to build GC by myself so that I can test your commit.

Another newb question: Does GC offer natively HRV data analysis ? What I mean is that if you were to offer the HRV data import from SML files, would GC be able to display HRV charts ?

Thank you

Frederic

[Ale Martinez]

El miércoles, 23 de septiembre de 2015, 16:14:23 (UTC-3), bardfr...@gmail.com escribió:

As I am still discovering GC, I am working on setting up my environment to build GC by myself so that I can test your commit.

Besides to play with the program Mark's videos in the main site and Joern's wiki are the best resources to learn about GC, IMHO.

To build a basic Linux version there is a document at https://github.com/GoldenCheetah/GoldenCheetah/blob/master/INSTALL-LINUX, the developer's guide in the main site explains the basics to work with git to hava an updated source.

Another newb question: Does GC offer natively HRV data analysis ? What I mean is that if you were to offer the HRV data import from SML files, would GC be able to display HRV charts ?

No, there is no HRV data analysis built in GC. 

[Jean Div]

Hey All

 

Another newb question: Does GC offer natively HRV data analysis ? What I mean is that if you were to offer the HRV data import from SML files, would GC be able to display HRV charts ?

No, there is no HRV data analysis built in GC. 

+1 - HRV could be quite nice. Is it stored in the file on a Garmin 920xt/Fenix3 .tcx?

My requests for running features are:

1. Critical Velocity chart to display run speed in Min/KM (currently km/h)

2. Add Min/KM to the 'ride' charts under the 'Running' section (currently contains Vertical Oscillation, GCT and Cadence).

3. Add 'Stride Length' to the 'ride' charts under the 'Running' section (Stride length = distance covered in meters/(cadence*2) - currently contains Vertical Oscillation, GCT and Cadence).

4. Perhaps include a calculation to forecast a Running Race of a particular distance based on the files at hand, as is currently done for 'Estimated V02Max' for cycling. This could probably come from VDOT tables. Race distance either customisable or predefined - 5km, 10km, Half Mar, Marathon. This would allow a user to track their progress towards a goal race.

5. Calculation for 'Estimated Athlete V02 Max' from running. Potentially use VDOTS.

6. Ability to customise the 'Peak Pace' Intervals for running - e.g. only track certain pre-defined times / distances.  e.g. 5m, 10m, 1hr as opposed to all by default. This is a bit broader - would be great for Cycling and Swimming too, but thought I'd mention here.

7. Deeper lap analysis functionality, as discussed elsewhere. Again, broader than just for running. The most common one that I use is a scatter chart of interval avg. HR vs. Interval Avg. xPace for the laps over time.

8. I know that Cycling has the GREAT function of recognising routes, but it would be great if we got an option for Running (or swimming, or cycling) when dragging an interval in for comparison that offered something along the lines of 'include similar intervals' where the software automatically brought in intervals of a similar HR/pace/distance/time/ascent (not 100% sure how to teach this, perhaps a pop-up tickbox?)

I hope that this is useful..?

Ta,

Jean

 

[roland....@gmail.com]

Hello Ale,

my name is Roland, I'm joining this coversation because Suunto *.sml support is something I would be interested as well:

The Suunto Movescount let's you export the data files as *.kml, *.gpx, *.xlsx, *.fit and *.tcx files and I could import the gpx, fit and tcx files into GC but unfortunately none of them showed all values correct:

The *.fit file had the average speed and power data wrong: Average Speed was 62.6 (should be around 32.8) and the average power was 116 Watts (should be around 235W). The Good thing about the fit File Import was that it imported the intervals correctly (I use the lap feature on the ambit to make intervals)

The gpx and gcx File imports showed avg. speed and avg. power correct but got the laps / intervals all wrong.

I'll attach the sml, fit, gpx and tcx files (all from the same ride) to this post so if you have time to look at it it might help you to figure out how to implemet sml support into GC. This specific ride contains 13 laps, every lap with an even number (2,4,6 etc.) is a "power interval", every uneven one is "relax interval" - therefore this ride had 6 power intervals, every one roughly about 3 minutes long.

The Suunto Ambit has a native Interval function as well but it's it kind of automatic and triggers on GPS position or GPS distance and therefore it is only usable for people who are training on a sports track. It doesn't work when you do your intervals on a public road where you have to consider villages, traffic lights etc, roundabouts etc. and therefore have to start your intervals manually whenever you have a stretch of road in front of you where you can power up uninterupted for the next 3 minutes. For this reason I'm using the lap function to track my intervals because I can start the laps manually on the Suunto Ambit.

The attached sml file is from a Suunto Ambit 2, it might look different from a Ambit 3 though (I don't own a Ambit 3)

regards, Roland

[Ale Martinez]

Hi Roland, thanks for the test files, I wanted to test an SML ride with power, with a minor correction (BikePower instead of Bikepower keyword) it seems to import reasonably, I've attached the result in native GC .json format so you can test in your current version, your feedback will be welcome!

[roland....@gmail.com]

Hello Ale,

thanks a lot for the File. Yes, looks good. The reason why I want to use something additional to Suunto Movescount to analyse and plan my training is because there seems to be an error in Movescount for the average power: For that particular ride the average power in Movescount is 283W which is far too high (my FTP is around 280Watts according to a Lactat step test), in Golden Cheetah it is now 225W which if more realistic. I read somewhere that in Movescount they don't take the coasting into consideration (therefore when you don't spin your crank and the power is zero) and because of that the average power values of the whole ride are far too high.

The numbering of the laps is offset by one in comparison to Movescount, for this particular ride in GC it is 0 to 12 while in Movescount it is 1 - 13. I think starting with 1 would make more sense...

I guess you will add the ability to import Suunto sml files into the next version, any idea when that version will be released ? Or is there a way where I could download a executeable for Win7 64bit that includes that functionality in the meantime ? I fear I'm not deep enough into computers to compile by myself....

regards, Roland

[Ale Martinez]

Roland, thanks for your feedback, next release is scheduled for end of the year but development builds are published in advance for testing purposes, so stay tunez to the announcements here in the forum.
Cheers, Ale.

[roland....@gmail.com]

Alejandro,

ok, thanks. I'll watch out for development builds.

regards, Roland

[Ale Martinez]

Hi Jean, thanks for your ideas, some update for the running related points follows.

WRT to HRV there had been discussions before, for example: https://groups.google.com/forum/#!searchin/golden-cheetah-users/HRV/golden-cheetah-users/iyX3Qso8U44/rx1DQuDbQj4J

El sábado, 26 de septiembre de 2015, 14:51:10 (UTC-3), Jean Div escribió:

1. Critical Velocity chart to display run speed in Min/KM (currently km/h)

The axis is speed in km/h but the tooltip has also pace in units according to the sport, for v3.3 I added distance (m&yd for swimming, km&miles otherwise) to allow looking for time&pace for distance in any of the curves to see what you have done and what the selected model estimates you can do.

The ability to change the axis could be interesting but it looks like a lot of work, IMHO.

 

2. Add Min/KM to the 'ride' charts under the 'Running' section (currently contains Vertical Oscillation, GCT and Cadence).

3. Add 'Stride Length' to the 'ride' charts under the 'Running' section (Stride length = distance covered in meters/(cadence*2) - currently contains Vertical Oscillation, GCT and Cadence).

For 2 apply the same as for 1 before, but additionally in v3.3 there exists the possibility to add user defined curves based on formulas, so the two above can be easily added:

Pace [min/km] -> 60/SPEED

Stride/Stroke Length [m]: SPEED*1000/60/CADENCE/2 

4. Perhaps include a calculation to forecast a Running Race of a particular distance based on the files at hand, as is currently done for 'Estimated V02Max' for cycling. This could probably come from VDOT tables. Race distance either customisable or predefined - 5km, 10km, Half Mar, Marathon. This would allow a user to track their progress towards a goal race.

5. Calculation for 'Estimated Athlete V02 Max' from running. Potentially use VDOTS.

Daniels VDOT is added as a new metric in v3.3, it is computed for each activity looking for the maximum value from 4 minutes to 4 hours so it can be used as Bests metrics and/or LTM charts to plot progress, also included in v3.3 is Daniels TPace derived from VDOT which roughly corresponds to the pace the athlete could sustain for near 1 hour (technically 90% of velocity at VDOT using Daniels/Gilbert formulas).

Predictions for standard distances can easily be looked based on VDOT, may be in 4.0 we could compute them. Alternatively one could use the VDOT calculator to look for the required VDOT (or TPace) for the target race and then track VDOT (or TPace).

6. Ability to customise the 'Peak Pace' Intervals for running - e.g. only track certain pre-defined times / distances.  e.g. 5m, 10m, 1hr as opposed to all by default. This is a bit broader - would be great for Cycling and Swimming too, but thought I'd mention here.

7. Deeper lap analysis functionality, as discussed elsewhere. Again, broader than just for running. The most common one that I use is a scatter chart of interval avg. HR vs. Interval Avg. xPace for the laps over time.

Yes, when these features become available for cycling we could adapt them for running&swimming.

 

8. I know that Cycling has the GREAT function of recognising routes, but it would be great if we got an option for Running (or swimming, or cycling) when dragging an interval in for comparison that offered something along the lines of 'include similar intervals' where the software automatically brought in intervals of a similar HR/pace/distance/time/ascent (not 100% sure how to teach this, perhaps a pop-up tickbox?)

Sorry, I don't understand this, how it needs to be different for running than for cycling? 

[bardfr...@gmail.com]

I am really appreciative of being able to share my feedback from GC and also submit my ideas for future releases, thank you GC Team !

Here is a list of features for running that I would like to be considered/implemented in GC :

- A thorough gear tracking. I understand that the filters can serve as such but I would find it very useful to have the ability to create "Gear" and assign info to each of them such as

Brand, Model, Price, Size, Date purchased, Pictures....etc)

- The ability to track injuries, weight, fat percentage...and other metrics on any given day. This is the notion of "Individuals or Athlete" that would be nice to have added

- Training Load chart to take into account TRIMP, CTL(Fitness), ATL(Fatigue), Performance (TSB)

- I know this one is already bing addressed : HRV tracking and analyzing for recovery and overtraining purposes

- More maps: As of today, we have the ability to display satellite or terrain maps from Google. It would be a nice addition to have map layers like OpenStreetMap, USGS...

- Elevation Correction : For a given run, the ability to correct the elevation by using the NASA SRTM data for example

- (Bonus) A "Route library" : A place to store routes that I plan on running on the future.

[Andrea Sabba]

Hi Ale, 

what about this feature https://groups.google.com/d/msg/golden-cheetah-users/Vv84ulMAz_Q/pYy0k5SoCwAJ?

Is it confirmed?

Thanks

Andrea

Il giorno sabato 5 settembre 2015 01:59:59 UTC+2, Ale Martinez ha scritto:

Now 3.2 is in the wild and the users have had some time to test the running features I think it would be a good time to think about the next versions, anyone who wish to contribute his ideas is welcome to comment here or to make a feature request at the tracker.

For my part I recently added an initial implementation of a little VDOT Calculator to master here is how it looks in metric and imperial units:

Even when the CP and W' Estimator in v3.2 can be used for running requires 2 relatively close tests and its use is not that common, Daniels' VDOT on the other hand can be used to estimate training paces (including T-Pace) using one race result and I find it very useful.

[Ale Martinez]

Yes, it will be configurable at athlete level

[Nathan Townsend]

Hi Guys,

To weigh in on this discussion.  Suunto HRV analysis uses a complex proprietary analysis developed in Finland.  Initially this analysis was available in Suunto software (can't remember the name) that came with the HRMs.  They then stopped developing this application and shifted over to Movescount.  Meanwhile, the team of developers of the HRV analysis methods (which uses neural networks and a whole bunch of complex time series maths) created a spinoff company.... you guessed it..... FirstBeat Technologies. Whilst FirstBeat own the rights to the analysis method (which derives breathing frequency and EPOC for example), I assume that Suunto made some sort of agreement to keep it in Movescount.  Since those algorithms are proprietary and FirstBeat own it, then obviously they will not give away their IP to an open source project.  So don't expect EPOC to appear in GC any time soon (or ever).

Regarding, a more basic HRV analysis, there is something very very simple that can be done.  The best work in this area has been developed over the past 10 yrs by Martin Buccheit and Paul Laursen and their PhD students Dan Plews and Jamie Stanley.  These guys have mainly zeroed in on a single variable which is lnRMSSD which can easily be calculated from a 5 min seated or supine early morning recording.  

Ale, if you're interested we can discuss this offline and I can assist with ideas regarding how this could be implemented and features.  Shoot me an email anytime!

[Ale Martinez]

Update on VDOT:

1) now we have equivalent times for distances from 1500 up to marathon on VDOT Calculator:

2) A new function can be used on LTM charts formulas to track VDOT equivalent times for distance:

vdottime(VDOT, distance)

Computes the equivalent time (in seconds) for distance (in km) at VDOT, tipically used with VDOT metric as first parameter which is computed from the best speed average from 1500 to marathon, for example:

vdottime(VDOT, 10)

computes the equivalent time for 10k based on best performance for the selected activities.

[mshalett]

Nathan,

was/is the Polar HRV analysis (rs800 series watches)also the same Finnish algorthm? 

Monte

[Rohan]

 With the new STRYD running power meter available can average power be displayed in the Activity Summary and in intervals? At present it only comes up under efforts

[Ale Martinez]

El domingo, 1 de noviembre de 2015, 4:07:39 (UTC-3), Rohan escribió:

 With the new STRYD running power meter available can average power be displayed in the Activity Summary and in intervals? At present it only comes up under efforts

Yes, to support STRYD we should enable all (or most) of the power derived metrics for running, for multisport athletes also it would be desirable to have independent Power Zones for Cycling and Running, similar to Pace Zones for Running and Swimming. Would be interesting to have some files for testing purposes.

[Rohan]

 It would be nice if we could graph this index of running efficiency: Speed / Watts/kg. If we use SI units it should come out at around 0.7.

 

 

[Ale Martinez]

El domingo, 1 de noviembre de 2015, 21:36:47 (UTC-3), Rohan escribió:

 It would be nice if we could graph this index of running efficiency: Speed / Watts/kg. If we use SI units it should come out at around 0.7.

In v3.3 you can use formulas to chart that figure in the Ride chart using sample values: SPEED/(POWER/Athlete_Weight)

and in LTM charts using average values: Average_Speed/(Average_Power/Athlete_Weight)

[Rohan]

That's great. The GC team truly is amazing.

I have only done three runs with my Stryd so far but it seems that my running efficiency deteriorates as my pace quickens.

6:14 min/km---> 3.4 W/kg -->0.79 (m/s)/(W/kg)
5:38 min/km --> 4.1 W/kg -->0.72 (m/s)/(W/kg)
5:09 min/km --> 4.5 W/kg -->0.719 (m/s)/(W/kg)
4:07 min/km --> 5.9 W/kg --> 0.69 (m/s)/(W/kg)

It may be interesting to be able to monitor efficiency at a set pace over a period of time. For instance, currently I run 40 pace at an efficiency of 0.69. Come January when I expect to be fitter, will this value have improved?

[Nathan Townsend]

Hi Rohan,  I have a spreadsheet already with calculations for "metabolic power" in running.  I'd be very interested to see how the power values from this Stryd device compare to the metabolic power calculations.  By the looks of it, the Stryd values are estimated external power, whereas, metabolic power is total internal energy expenditure expressed in watts.  Therefore, if we have both sets of data, we can express the ratio which of course should correspond to efficiency.  

Any chance you can send me a raw datafile of an interval type workout on consistent flat terrain?

[Rohan]

Nathan,

See if these are flat enough for you. Perhaps some sections may be.

Rohan

[Nathan Townsend]

Nice one.  I'll have a play around with these and see what comes of it.  The only problem that I can see straight off the cuff is that only velocity is measured, so I'll have to derive acceleration which usually comes which adds noise that will then need to be filtered out.

Can anyone suggest an appropriate filter for smoothing of acceleration data?  

[Nathan Townsend]

On Tuesday, 3 November 2015 08:28:43 UTC+3, Rohan wrote:

Nathan,

See if these are flat enough for you. Perhaps some sections may be.

Rohan

Ok so I've written to the company. Hopefully they'll send me a couple of freebies.  In the meantime, if you're interested Rohan, do you want to run a little case study? (excuse the pun).

What I would need you to do is conduct a critical power test.  This would involve two separate sessions in which you do a 2 max effort time trials in each session as follows:

day 1:  10min TT -> 30-40min recovery -> 2min TT

day 2:  14min TT -> 30-40min reco  ->  6min TT

IMO the best way to perform these TTs to get the best possible picture of max performance, is to make them quasi time to exhuastion.  ie: you treat the effort like a max effort 6 min TT (or whatever duration), and in particular the last minute must be extremely HARD.... maximal effort in fact.  As you approach the last 20 sec you must give it everything you've got, basically its like a sprint to the line, however, it is not imperative to stop right on 6 min.  You should only stop right on 6min if you are sprinting maximally but it is noticable that you are actually slowing down due to fatigue. If you feel like you paced the effort slightly incorrect and you have a little left in the tank and can keep going for another 20-30sec, then KEEP GOING AT MAX EFFORT UNTIL YOU START SLOWING DOWN.

Same goes for less than the target duration ie: lets say you lift the pace with one minute to go, you're giving it everything, but you went too early and you're slowing down with 20 sec left.  You can stop early as long as the effort was truly maximal and you can't even anymore.  

So think of the target durations as +/- 20 sec  EXCEPT for the 2min TT.  That should never be less than 2min.  

Lets see what happens with Phil's work-balance model once we have this rolling eh?

 

[Nathan Townsend]

On Tuesday, 3 November 2015 16:19:58 UTC+3, Nathan Townsend wrote:

On Tuesday, 3 November 2015 08:28:43 UTC+3, Rohan wrote:

Nathan,

See if these are flat enough for you. Perhaps some sections may be.

Rohan

 good training sets these anyway.  You'll love em :-)

[Ale Martinez]

Now in current master power derived metrics are enabled for non-cycling activities when power data is present, here is the summary using one of your contributed files:

[Nathan Townsend]

On Tuesday, 3 November 2015 17:53:06 UTC+3, Ale Martinez wrote:

Now in current master power derived metrics are enabled for non-cycling activities when power data is present, here is the summary using one of your contributed files:

My initial thoughts when I reviewed these files earlier is that the power is high.  For example lap 4 interval on Nov 1 is 397W for just over 20min, but the pace is about 4:05 km/min which demands low 50s for VO2 (ml/kg) in a good runner with average economy.  At 80kg this is in the range 4.0 - 4.25 l/min.  To sustain 400W though in cycling requires closer to 4.9 L/min.  If Rohan was approx 95 kg then it would probably be about right.  

[Rohan]

They seem very high to me too as I'm only 68 kilograms and I have only been running for six weeks. Prior to this my cycling FTP was set at 250 although it was at 285 earlier in the year.

[Rohan]

I can do that but not for a few weeks. Being only six weeks into training it has become very obvious that running fast is the big risk when it comes to injury so I'm only doing one fast training session per week, plus I may have a duathlon on this Sunday and 5km fun run the following Sunday. I was also hoping to do a 1km time trial which in addition to the 5km time trial should allow me to calculate CP &  W'.

Would the results from the duathlon be of any use? I'm not doing the ride (someone has to watch the kids) so it would be two 3000m runs with around an hour to rest between them.

[Nathan Townsend]

On Wednesday, 4 November 2015 05:22:33 UTC+3, Rohan wrote:

They seem very high to me too as I'm only 68 kilograms and I have only been running for six weeks. Prior to this my cycling FTP was set at 250 although it was at 285 earlier in the year.

Ok yeah way too high then. In the accompanying app is there somewhere that you tell the device your weight?  I wonder if that might correct the values downwards to something more realistic.

No hurry regarding the test efforts.  If I'm lucky then maybe the company will send me a freebie to trial some validation work.  We have a fancy instrumented treadmill with force platforms, so we can measure the actual power in running.

I think this is important for people to understand marketing hyperbole.  This device is not a power "meter" because a meter measures something directly eg: SRM measures  force and cadence directly.  This device does not measure force directly.  I expect that it uses complex pattern recognition signal processing to estimate ground reaction force from the accelerometry data.

[Rohan]

Looking at that 5km test of 20:32 (397 watts) gives an estimated VO2max of 48.3 ml/kg/min or 3.3 litres per minute (http://www.runnersweb.com/running/vo2_js.html). If in cycling 400W takes closer to 4.9 L/min, that just happens to be a 50% difference which also happens to be the same between my CPr (Stryd) and my CPc(SRM).

At this stage I'm not placing much credence on the device's accuracy but it does seem to do everything else that I'd expect it to do and it is allowing me to train with power which is something I far more comfortable doing than trying to estimate paces over varying terrain.

The Stryd does all of its calculations in w/kg. You enter your weight and it gives you a power reading. The default weight is 70 kg and I had entered my weight setting anyway.

[Ale Martinez]

I found interesting to compare the NP computed from Stryd power with the LNP computed according to the GOVSS algorithm for long intervals since both use the same normalization (30s MA + 4th power weighting). Using 68kg as athlete weigth LNP seems to be around 8-10% lower -on average- than NP from Stryd power:

[Mark Liversedge]

On Wednesday, 4 November 2015 23:25:15 UTC, Ale Martinez wrote:

I found interesting to compare the NP computed from Stryd power with the LNP computed according to the GOVSS algorithm for long intervals since both use the same normalization (30s MA + 4th power weighting). Using 68kg as athlete weigth LNP seems to be around 8-10% lower -on average- than NP from Stryd power:

What kind of correlation are you seeing?

Be interesting to see if they track or where they diverge...

Mark 

[Ale Martinez]

Mark, I have only the two activities posted in this thread with Stryd power data, so R2=1 :-)

More seriously, dividing them in 5' segments this is what I'm seeing with 22 data points:

[Mark Liversedge]

On Thursday, 5 November 2015 12:07:11 UTC, Ale Martinez wrote:

Mark, I have only the two activities posted in this thread with Stryd power data, so R2=1 :-)

More seriously, dividing them in 5' segments this is what I'm seeing with 22 data points:

Interesting .. how to decide if one is undercooked or the other is overcooked !?

:)

Mark 

[Michele Cea]

Hi Ale,

a quick question...what about the opportunity to read .fit files with swim work out. Including lenght splits....it will makes life much easier for triathletes :-)

Is that going to be part of the 3.3 or 4.0 version?

thank you very much

Michele

Il giorno sabato 5 settembre 2015 01:59:59 UTC+2, Ale Martinez ha scritto:

Now 3.2 is in the wild and the users have had some time to test the running features I think it would be a good time to think about the next versions, anyone who wish to contribute his ideas is welcome to comment here or to make a feature request at the tracker.

For my part I recently added an initial implementation of a little VDOT Calculator to master here is how it looks in metric and imperial units:

Even when the CP and W' Estimator in v3.2 can be used for running requires 2 relatively close tests and its use is not that common, Daniels' VDOT on the other hand can be used to estimate training paces (including T-Pace) using one race result and I find it very useful.

[Ale Martinez]

El miércoles, 11 de noviembre de 2015, 11:08:18 (UTC-3), Michele Cea escribió:

Hi Ale,

a quick question...what about the opportunity to read .fit files with swim work out. Including lenght splits....it will makes life much easier for triathletes :-)

Is that going to be part of the 3.3 or 4.0 version?

I don't know, this was the last update on the issue: https://github.com/GoldenCheetah/GoldenCheetah/pull/1409 

In v3.3 there are no new swimming features that I'm aware, only some minor bug fixes.

[Mark Liversedge]

We can revisit this in v4.0.

v3.3 is entering the release cycle this weekend.

Mark 

[Ale Martinez]

Ok, I did the attempt and here is the result: https://github.com/GoldenCheetah/GoldenCheetah/pull/1659, It is not fully tested since I don't own a device capable to generate this kind of files, but perhaps it could be added to RC1 for users to test since it shouldn't affect the import of non lap swimming activities.

[Steven]

Hi Ale,

Can you help me about a running chart? I have only running files, and I can see at the tarining RR, CTL, TSB, pace, etc values. But I can not make a reguular PMC or Skiba stress chart, beacuse it's always empty. I think it's some banal error I've made, but can not figure out what is it. Because at cycling data it's works correcty. 

Thanks!

[Ale Martinez]

El lunes, 14 de diciembre de 2015, 5:11:55 (UTC-3), Steven escribió:

Can you help me about a running chart? I have only running files, and I can see at the tarining RR, CTL, TSB, pace, etc values. But I can not make a reguular PMC or Skiba stress chart, beacuse it's always empty. I think it's some banal error I've made, but can not figure out what is it. Because at cycling data it's works correcty.  

Hi Steven, PMC data on summary is based on TriScore for running activities, the PMC (TriScore) preset chart is what you are looking for

[Steven]

Thanks!

[JuAucout]

Hi Ale,

I am a recent user of Golden Cheetah v3.3 and started importing my data from Garmin Connect. I am using a Garmin 620 watch for running and can't find a way to get the vertical oscillation data, or stride length data.
I guess that it is possible as I can create this plot. Is there a tutorial or a way to see these running metrics ?

Thanks a lot!

Ju




Le mercredi 21 octobre 2015 21:44:16 UTC+2, Ale Martinez a écrit :

Hi Jean, thanks for your ideas, some update for the running related points follows.

WRT to HRV there had been discussions before, for example: https://groups.google.com/forum/#!searchin/golden-cheetah-users/HRV/golden-cheetah-users/iyX3Qso8U44/rx1DQuDbQj4J

El sábado, 26 de septiembre de 2015, 14:51:10 (UTC-3), Jean Div escribió:

1. Critical Velocity chart to display run speed in Min/KM (currently km/h)

The axis is speed in km/h but the tooltip has also pace in units according to the sport, for v3.3 I added distance (m&yd for swimming, km&miles otherwise) to allow looking for time&pace for distance in any of the curves to see what you have done and what the selected model estimates you can do.

The ability to change the axis could be interesting but it looks like a lot of work, IMHO.

 

2. Add Min/KM to the 'ride' charts under the 'Running' section (currently contains Vertical Oscillation, GCT and Cadence).

3. Add 'Stride Length' to the 'ride' charts under the 'Running' section (Stride length = distance covered in meters/(cadence*2) - currently contains Vertical Oscillation, GCT and Cadence).

For 2 apply the same as for 1 before, but additionally in v3.3 there exists the possibility to add user defined curves based on formulas, so the two above can be easily added:

Pace [min/km] -> 60/SPEED

Stride/Stroke Length [m]: SPEED*1000/60/CADENCE/2 

4. Perhaps include a calculation to forecast a Running Race of a particular distance based on the files at hand, as is currently done for 'Estimated V02Max' for cycling. This could probably come from VDOT tables. Race distance either customisable or predefined - 5km, 10km, Half Mar, Marathon. This would allow a user to track their progress towards a goal race.

5. Calculation for 'Estimated Athlete V02 Max' from running. Potentially use VDOTS.

Daniels VDOT is added as a new metric in v3.3, it is computed for each activity looking for the maximum value from 4 minutes to 4 hours so it can be used as Bests metrics and/or LTM charts to plot progress, also included in v3.3 is Daniels TPace derived from VDOT which roughly corresponds to the pace the athlete could sustain for near 1 hour (technically 90% of velocity at VDOT using Daniels/Gilbert formulas).

Predictions for standard distances can easily be looked based on VDOT, may be in 4.0 we could compute them. Alternatively one could use the VDOT calculator to look for the required VDOT (or TPace) for the target race and then track VDOT (or TPace).

6. Ability to customise the 'Peak Pace' Intervals for running - e.g. only track certain pre-defined times / distances.  e.g. 5m, 10m, 1hr as opposed to all by default. This is a bit broader - would be great for Cycling and Swimming too, but thought I'd mention here.

7. Deeper lap analysis functionality, as discussed elsewhere. Again, broader than just for running. The most common one that I use is a scatter chart of interval avg. HR vs. Interval Avg. xPace for the laps over time.

Yes, when these features become available for cycling we could adapt them for running&swimming.

 

8. I know that Cycling has the GREAT function of recognising routes, but it would be great if we got an option for Running (or swimming, or cycling) when dragging an interval in for comparison that offered something along the lines of 'include similar intervals' where the software automatically brought in intervals of a similar HR/pace/distance/time/ascent (not 100% sure how to teach this, perhaps a pop-up tickbox?)

Sorry, I don't understand this, how it needs to be different for running than for cycling? 

[Ale Martinez]

El martes, 22 de marzo de 2016, 11:58:56 (UTC-3), JuAucout escribió:

Hi Ale,

I am a recent user of Golden Cheetah v3.3 and started importing my data from Garmin Connect. I am using a Garmin 620 watch for running and can't find a way to get the vertical oscillation data, or stride length data.

To have Vertical Oscillation and Ground Contact Time you need to import the original .fit file since the .tcx format doesn't carry these data. Stride length is not directly supported byt it can be computed from speed and cadence.

 

I guess that it is possible as I can create this plot. Is there a tutorial or a way to see these running metrics ?

Yes, this plot should work once you have the data imported, check for it on Ride and/or Edit tabs, and no, there is no tutorial that I know. 

[Jorge Gonzalez]

Hi Ale Martinez

I am new to golden cheetah and currently loading bike and run data. In the summary page, run data for pace shows in min/km. I would like to change it to display in min/mile. Is this possible? I set the units to imperial for the user but this seems to have no effect.

Thanks for your help!

[Ale Martinez]

Yes, there is a metric pace checkbox on pace zones config